Table 2.

Extraction table of the evaluated studies.

Author	Sample	Study Design	Exercise Temporality	Exercise Protocol	Memory Assessment	Results
Stones et al. (1993) [62]	20 older adults, Mage = 84.5	Experimental; between-subject	Memory battered occurred before exercise, immediately after exercise and then 30-min post-exercise	15-min exercises, which occurred while sitting in a chair (e.g., stretching, low-intensity aerobic activity, slow rhythmical movement)	Word fluency	Exercise was associated with greater semantically cued memory (p < 0.01).
Schramke et al. (1997) [63]	Two age groups, each including 48 adults. Younger group, 18–38 year. Older group, 60–80 year.	Experimental; within-subject	Exercise occurred either at rest or during encoding, and similarly, either during retrieval or not.	5–7 minutes of walking in a long internal corridor.	CVLT; California verbal learning test	There was no difference in learning that was due to initial exercise condition, but both age groups showed greater recall when state was congruent before learning and delayed recall.
Labban et al. (2011) [6]	48 young adults (Mage = 22.0)	Experimental; between-subject	Exercise occurred before and after encoding	30-min of cycle ergometer exercise, with 20-min at moderate-intensity	Paragraph recall, with participants listening to two paragraphs and then recalling as much information as possible from the paragraphs	Exercise occurring prior to the memory task was effective in enhancing memory (p < 0.05).
Salas et al. (2011) [64]	80 college undergraduate students (46 women). Mage = 19.3, SD = 2.3	Experimental; between-subject factorial design. A 2 (encoding condition: walking vs. sitting) × 2 (retrieval condition: walking vs. sitting).	Exercise occurred either at rest or during encoding, and similarly, either during retrieval or not.	10 minutes of walking outside at a brisk pace	Word-list memory task (10 nouns presented sequentially for 6 s each)	Students who walked before encoding had significantly higher recall (M = 0.45, SD = 0.17) compared to students who sat before encoding (M = 0.36, SD = 0.15), F(1,76) = 6.34, ƞ2p = 0.08.
Nanda et al. (2013) [65]	10 healthy adult male medical students. Mage = 19.5, SD = 0.9	Quasi-experimental; within-subject	Exercise occurred between pre- and post- memory assessments.	Cycle ergometer exercise for 30-min at moderate-intensity of 70% of heart rate reserve	Spatial span and paired associates memory task	Spatial span did not increase from pre- to post, but paired associates was significantly higher after the exercise bout.
Schmidt-Kassow et al. (2014) [10]	49 right-handed German young adults (18–30 year)	Experimental; within-subject	Exercised during encoding	Self-selected walking pace during memory encoding	40-item (Polish) word list.	Experiment 1: words recalled during walking was higher than non-walking (5.5, SD = 3.3; vs. 4.8, SD = 4.2), F = 6.98, p = 0.02, ƞ2p = 0.31. Experiment 2: words recalled during walking was higher than non-walking (5.3, SD = 4.6; vs. 4.1, SD = 3.5), F = 6.44, p = 0.02, ƞ2p = 0.19.
Weinberg et al. (2014) [66]	23 participants (Mage = 20.6 year) in the exercise group and 23 (Mage = 20.2 year) in the control group.	Experimental; between-subject	Exercised during early consolidation	Isokinetic dynamometer knee extension exercise. Session consisted of submaximal voluntary dynamic contractions for a warm-up, maximal voluntary isometric contractions, and 6 sets of 10 repetitions of maximal voluntary knee extension contractions. Both legs were exercised. In the control (passive) group, the experimenter passively moved the participant leg between extension and flexion.	180 images from the IAPS. Follow-up memory recall assessment took place 48-h later. The retrieval task included 90 studied images and 90 new images. Participants were instructed to indicate “remember”, “familiar”, or “new” after seeing each image.	There was no valence × group interaction effect. There was a main effect for valence in that participants remembered more positive and negative images than neutral images.
Basso et al. (2015) [67]	85 young adults, Mage = 22.1	Experimental; between-subject	Memory tasks occurred before exercise and at various time-points after exercise (30–120 min)	50-min of vigorous-intensity exercise on cycle ergometer	Hopkins verbal learning test revised, modified Benton visual retention test, Digit span	Acute exercise improved prefrontal-cortex, but not hippocampal-dependent memory function.
Loprinzi et al. (2015) [68]	87 young adults, Mage = 21.4 year	Experimental; between-subject	Exercise before memory task	Light, moderate, and vigorous exercise	Spatial span and paired associates	Acute exercise was not associated with either memory outcome.
Bantoft et al. (2016) [69]	45 undergraduate students, Mage = 22.6 year (6.2)	Experimental; within-subject	Sitting, standing or walking during memory task	Low-intensity walking	Digit span	There were no differences in memory performance across the three conditions.
van Dongen et al. (2016) [19]	72 young adults, approximately 22 years	Experimental; between-subject	Exercise immediately after encoding and 4 hours after encoding	35 min of intermittent high-intensity exercise on cycle ergometer	Paired associates learning task	Exercising 4 hours after memory encoding was advantageous in improving memory function.
Crush et al. (2017) [70]	352 participants, mean age approximately 21 years	Experimental; between-subject	Exercise occurring before memory assessment	16 total groups, with groups ranging from 10 min of exercise to 60 min of exercise, including resting periods of either 5, 15, or 30 min	Spatial span	Shorter exercise recovery periods had a greater effect on memory performance.
Frith et al. (2017) [31]	88 participants (22 per group), approximate age = 21 years.	Experimental; between-subject	Exercise occurring before, during, and after memory encoding	15-min treadmill bout of progressive high-intensity aerobic exercise	RAVLT	High-intensity exercise prior to memory encoding was effective in enhancing long-term memory, for both 20-min delay (F = 3.36, p = 0.02, ƞ2p = 0.11) and 24-h delay (F =2.80, p = 0.04, ƞ2p = 0.09).
Keyan et al. (2017) [71]	49 undergraduates between 18–29 years	Experimental; between-subject	Exercise occurred during the early memory consolidation period	Stepping exercise for 10-min on a 15 cm stepper, with a goal of exercising at 50%–85% of max.	Viewed a film depicting a car accident. Involves 10 min of live footage depicting emergency workers attending the scene of a motor vehicle accident.	Exercise (vs. control) did not induce more recall of central (t = 0.11, p > 0.05) or peripheral (t = 0.42, p > 0.05) details of the accident film. However, those that exercise recalled more intrusive memories of the car accident (t = 2.36, p = 0.02, d = 0.68).
Keyan et al. (2017) [72]	54 healthy undergraduate students, Mage = 19.5 (3.0)	Experimental; between-subject	During a memory reconsolidation paradigm, participants either exercised or did not exercise after memory reactivation	20–25 min of incremental cycling	Trauma film depicting the aftermath of a highway car crash	The exercise with reactivation condition recalled more central details of the trauma film.
McNerney et al. (2017) [73]	Experiment 1: 136 young adults, Mage = 19.2 (1.2) Experiment 2: 132 young adults, Mage = 19.1 (1.2)	Experimental; between-subject	Exercise occurring before and after memory encoding	2-min of sprints	Paired associate learning, procedural learning, and text memory	Improvements in procedural and situation model memory occurred, regardless of whether exercise occurred before or after memory encoding.
Most et al. (2017) [74]	Experiment 1: 82 undergraduate psychology students (Mage = 19.9). Experiment 2: 83 undergraduate psychology students (Mage = 19.9). Experiment 3: 48 undergraduate psychology students (Mage = 19.2). Experiment 4: 75 undergraduate psychology students (Mage = 21.1).	Experimental; between-subject	Exercise occurring after memory encoding	5-min of step exercise	Paired faces and names.	Acute exercise in the early consolidation period enhanced memory.
Sng et al. (2017) [32]	88 participants, approximately 21–25 years (mean for each group)	Experimental; between-subject	Exercise occurred before, during and immediately after memory encoding	15-min moderate intensity brisk walking (self-selected)	RAVLT	Exercising before memory encoding was superior for enhancing learning (p = 0.05), 24-h memory recognition (p = 0.05) and 24-h memory attribution (p = 0.006).
Delancey et al. (2018) [40]	40 participants, approximately 20 years of age	Experimental; between-subject	Exercise occurring 4 hours after memory encoding	High-intensity bout of exercise for 15 minutes	RAVLT	Those who exercise during the consolidation period have a greater 24-h follow-up memory attribution (p = 0.04).
Haynes et al. (2018) [33]	24 participants (Mage = 20.9; SD = 1.9), with 66.7% being female.	Experimental; within-subject	Exercise occurring before, during, and after memory encoding	Self-selected brisk walking pace for 15-min	RAVLT	Short-term memory was greater in the visit that involved exercise prior to the memory task (F= 3.76, p = 0.01, ƞ2p = 0.79). Similar results occurred for long-term memory, but there were no exercise effects on prospective memory.
Labban et al. (2018) [75]	15 Participants; Mage = 22.7, SD = 3.1	Experimental; within-subject	Exercise occurring both before and after memory encoding.	30-min of moderate intensity cycling	RAVLT	Exercise that occurred before encoding (vs. control) was advantageous in enhancing long-term memory, including both 60-min delayed memory (p = 0.03) and 24-h delayed recall (p = 0.03).
Siddiqui et al. (2018) [76]	20 participants (60% male). Mage = 21.1; SD = 1.0	Experimental; within-subject	Exercise occurring both before and during memory encoding.	20-min treadmill walk at a self-selected brisk walking pace	The Deese-Roediger-McDermott (DRM) paradigm. Included a 15-item word list.	For both short-term and long-term memory, the visit the involved exercise before the memory task resulted in the greatest memory performance (F = 11.56, p < 0.001, ƞ2p = 0.38)
Wade et al. (2018) [77]	34 female participants; Mage = 20.5 (1.2) in the exercise group and 20.8 (1.8) in the control group.	Experimental; between-subject	Exercise occurred before memory encoding	15-min treadmill walk at a self-selected brisk walking pace	Emotional memory assessment using images from the IAPS (International Affective Picture System).	There were no statistically significant group differences across any of the assessment periods (i.e., 1-day, 7-day, and 14-day follow-up assessments).
Yanes et al. (2018) [78]	40 participants, Mage = 21.0	Experimental; between-subject	Exercise occurred before memory encoding	15-min treadmill walk at a self-selected brisk walking pace	6-paragraph passage for memory recall	Exercise before encoding had greater scores on the short-term and long-term memory assessments, but this did not reach statistical significance (F = 1.0, p = 0.32, ƞ2p = 0.03).
Zuniga et al. (2018) [79]	Experiment 1 (N = 30), Mage = 20.4 (1.8); Experiment 2 (N = 57), Mage = 20.6 (4.1) in low-fit group and Mage = 19.4 (1.6) in high-fit group.	Experimental; within-subject	Exercise occurred before memory encoding	3-min warm-up period on the treadmill, followed by 10-min of walking at either light or moderate-intensity.	Three lists of 30 concrete English nouns from the MRC Psycholinguistic database.	Both light-intensity (t = 2.79, p = 0.01) and moderate-intensity (t = 3.02, p = 0.006) recalled more words than the sedentary condition. Results were similar when comparing high-fit to low-fit individuals.

CVLT; California verbal learning test; IAPS, International Affective Picture System; RAVLT, Rey Auditory Verbal Learning Task.