§ Claim under review · Study
"In a human experiment (published in Current Biology), participants who exercised 4 hours after learning picture-location associations remembered more when tested 48 hours later than those who did not exercise."
Verdict
Accurate
Confidence
HighSummary
This claim checks out. A real 2016 study in the journal Current Biology tested 72 people who learned 90 picture and location pairings, then randomly assigned them to cycle immediately, cycle four hours later, or not exercise at all. When tested 48 hours later, the group that exercised four hours after learning retained more than the no-exercise group, while exercising immediately made no difference. One caveat: the post's graphic prints the wrong citation, naming "Alves et al." and the wrong page numbers, though the caption gives the correct reference. Another caveat: this is a single study of 72 people, the difference against the control group was statistically borderline, and the researchers did not measure why it happened. The broader research on exercise and memory timing is mixed, so the four-hour window should not be treated as a proven rule. To its credit, the post itself says the ideal timing and intensity are not yet established.
The readings
key figures from the evidencepost hoc pairwise comparison, delayed-exercise vs no-exercise
Why this verdict
Evidence
The study exists, is human, is randomized, and reports what the post says it reports. Seventy-two study participants learned 90 picture-location associations over a period of approximately 40 minutes before being randomly assigned to one of three groups: one group performed exercise immediately, the second performed exercise four hours later, and the third did not perform any exercise.
Forty-eight hours later, participants returned for a cued-recall test in a magnetic resonance scanner. The authors found that performing exercise 4 hr, but not immediately, after encoding improved the retention of picture-location associations compared to the no-exercise control group.
Performing exercise after a delay was also associated with increased hippocampal pattern similarity for correct responses during delayed retrieval.
The underlying numbers from the paper itself: memory retention was significantly different between groups (mNE = 0.800, mIE = 0.795, mDE = 0.866; FGroup = 4.83, pGroup = 0.011). Post hoc pairwise comparisons showed that retention in the DE group was higher than in the IE and NE groups, with no difference between the latter two groups (DE-IE: pTukeyHSD = 0.031; DE-NE: pTukeyHSD = 0.045; IE-NE: pTukeyHSD = 0.986).
Findings
✓ What's accurate 6
- The study is real, human, randomized, and published in Current Biology in 2016 by van Dongen and colleagues.
- Participants did learn picture-location associations, and the delayed group did exercise approximately four hours after learning.
- Memory was tested 48 hours later, and the delayed-exercise group did show higher retention than the no-exercise group.
- The post's additional statement that exercising immediately after learning had no effect matches the paper.
- The post's statement that brain imaging showed differences associated with successful retrieval matches the reported hippocampal pattern similarity finding.
- The post's own disclaimer, that ideal timing, intensity and type of exercise are not established, is consistent with the state of the literature.
≈ What's misleading 3
- Attribution error in the graphic. The infographic cites "Alves et al., Current Biology. 2016;26(6):676–681." The actual paper is van Dongen et al., Current Biology 2016;26(13):1722-1727. The author name, issue and page range in the image are all wrong. The caption gives the correct citation, so the post contradicts itself. This is a sourcing error rather than a factual distortion of the finding.
- Slight measure simplification. "Remembered more" describes a retention ratio (0.866 versus 0.800 relative to each group's own baseline), not simply a larger raw number of items recalled. The direction is correct but the metric is more specific than the phrasing implies.
- Scope. The claim is drawn from one study of 72 people with a borderline pairwise result against the control group (p = 0.045). The post presents this as a single experiment, which is honest, but a reader could take it as an established effect.
? What's uncertain 4
- Whether the "Alves et al., Current Biology 2016;26(6):676–681" reference corresponds to any real article could not be resolved. My search budget was exhausted before this was settled, so I cannot state whether it is a mistyped citation or a nonexistent one.
- Whether the four-hour finding has been directly and independently replicated. I found the effect cited approvingly in later reviews but no direct replication attempt, and I found evidence that results across the wider acute-exercise literature are split.
- The mechanism. The catecholamine explanation was inferred from animal work and was not measured in this study.
- Generalization beyond healthy young adults, beyond a spatial associative task, and beyond 35 minutes of cycling at up to 80% of maximum heart rate is not supported by this study.
Sources
6 of 6 linked to recordsvan Dongen EV, Kersten IHP, Wagner IC, Morris RGM, Fernández G. "Physical Exercise Performed Four Hours after Learning Improves Memory Retention and Increases Hippocampal Pattern Similarity during Retrieval." Current Biology 2016;26(13):1722-1727. doi:10.1016/j.cub.2016.04.071
University of Edinburgh Research Explorer record (co-author Morris institutional listing)
Cell Press press release via ScienceDaily and Neuroscience News, June 16 2016
Sci.News summary quoting the methods section verbatim
Herold et al., Frontiers in Aging Neuroscience 2021, contextual citation of the finding within the wider literature
Loprinzi et al., "The Temporal Effects of Acute Exercise on Episodic Memory Function," Brain Sciences 2019 (systematic review with meta-analysis)