Can Sleep Improve Memory in Dogs?
Hungarian researchers studied whether sleep has the same memory consolidation effect in dogs as it does in humans.
We’ve all heard the familiar advice about getting a good night's sleep before a test. Such advice is rooted in the theory that sleep functions as a mode of memory consolidation,1 which is the conversion of short-term memory into long-term memory.2 According to 1 study, when we sleep, our brains refresh memory circuits—an activity that is incompatible with the processing of sensory information that occurs while we’re awake.3
Could this same memory consolidation take place in dogs when they sleep? The results of a study conducted by a team of researchers in Hungary recently revealed a connection between sleep and learning in dogs, providing novel evidence of sleep-dependent memory consolidation in man’s best friend.
To date, most studies evaluating the link between sleep and memory consolidation have been performed in humans and laboratory rodents. Sleep studies in dogs have focused primarily on brain activity with neurologic conditions such as epilepsy.4 According to the current study’s researchers, canine studies on sleep function could provide useful insight into the function of human sleep, given dogs’ long domestication his- tory and humanlike social skills.5
For dogs and other mammals, the sleep cycle consists of 3 main phases: nonrapid eye movement (non-REM) sleep, REM sleep, and wakefulness.6 Electroencephalogram (EEG) studies have demonstrated the cyclic brain wave activity throughout the sleep cycle (FIGURE):
- Beta waves: alert wakefulness
- Alpha waves: drowsy wakefulness
- Theta waves: REM sleep
- Delta waves: non-REM sleep
For the current study, the research team conducted a 2-part learning experiment to evaluate sleep and memory consolidation in dogs.
Part 1: The Effect of Learning on Sleep Physiology
The researchers first looked at how learning an unfamiliar verbal command affects brain wave activity. Fifteen adult pet dogs were selected based on their ability to follow verbal “sit” and “lie down” commands and to understand the hand signals for each command. Each dog participated in command learning (CL) and nonlearning (NL) scenarios in the afternoons on different days.
For the CL scenario, dogs first had a teaching session during which they learned unfamiliar English verbal commands for “sit” and “lie down.” Next, they had a baseline testing session of the English commands. Immediately following that session, the dogs underwent 3-hour noninvasive polysomnography and were then retested with the English commands after awakening.
For the NL scenario, teaching and baseline sessions consisted only of familiar Hungarian commands for “sit” and “lie down,” immediately followed by 3-hour polysomnography. Dogs performed the same number of “sit” and “lie down” actions and received the same amount of food reinforcement in both scenarios; the only difference between the scenarios was the learning of new information or the lack thereof. Dogs were not retested in this scenario.
From the polysomnography recordings, researchers collected sleep structure data, including sleep duration and non-REM and REM durations. EEG spectral analysis was performed to evaluate the relative contribution of each wave frequency (beta, alpha, theta, delta) to the overall brain wave pattern. For only the CL scenario, behavioral data were collected to compare performance between the baseline and retest sessions and to evaluate the relationship between performance and brain activity.
Contrary to the researchers’ expectations, but similar to findings in some human sleep studies,7 learning had no influence on the dogs’ sleep structure. EEG spectral analysis indicated a few notable changes in postlearning brain wave activity during REM and non-REM sleep. In REM sleep, theta activity significantly increased; in non-REM sleep, delta activity significantly increased and alpha activity significantly decreased. Within non-REM and REM sleep, slow-wave activity changes were inversely related to fast-wave activity changes. This finding, coupled with the decreased alpha activity in non-REM sleep, may have indicated deeper sleep after learning,8 the researchers noted.
In the CL scenario, performance significantly improved from baseline to retest, suggesting a positive effect of sleep on learning. Decreased delta activity and increased beta activity during REM sleep were significantly correlated with higher performance, whereas wave activity during non-REM sleep and theta and alpha activity during REM sleep did not correlate with performance.
Anna Kis, PhD, the study’s lead researcher, said via email that part 1 results indicate that the “individual sleep EEG spectrum is related to learning performance upon awakening, suggesting that differences in sleep pattern and brain activity during sleep have an effect on postsleep behavior.”
Part 2: The Effect of Sleep and Awake Activity on Learning
The researchers then evaluated how various postlearning activities affect learning. Fifty-three adult pet dogs underwent a similar CL scenario as in part 1. After that, instead of undergoing polysomnography, they participated in 1 of 4 activities during a 1-hour retention interval (RI):
- Playing with a Kong toy
- Learning a new command
- Sleeping in their owner’s car
- Leash walking around the university campus
Researchers did not include a “resting” awake RI activity because keeping a dog awake while resting could stress the animal, which would not only raise animal welfare concerns but also negatively affect memory.9 After the RI, dogs were retested with the English commands and had an obedience session with the Hungarian commands. One week later, dogs were tested again with the English commands to evaluate their long-term memory. The percentage of correct actions was calculated for all sessions (baseline, retest, obedience, long-term). Researchers compared performances among all sessions and used regression analysis to determine relationships between performance, test session (baseline, retest, long-term), and RI condition.
Overall, performance was significantly related to test session, “suggesting that differential learning patterns emerged as a consequence of the various activities following the initial learning task,” the researchers wrote. Interestingly, performance did not significantly improve between baseline and retest for the sleep condition; this was likely due to the short RI duration. However, performance significantly improved at the long-term test session for the sleep, walk, and play conditions, suggesting that at-home night sleep after learning improves memory consolidation under certain conditions.
Performance remained relatively unchanged between test sessions for the learning condition, indicating that learning a new command can interfere with memory consolidation of the original command. For the play condition, performance significantly decreased between baseline and retest, potentially due to the emotional arousal of playtime.
More Research and Real-World Applications
According to the researchers, this study’s novel findings open the door for future research and real-world application of the link between sleep and learning in dogs. In the research realm, studies could examine whether age-related changes in sleeping, brain activity, and memory function influence memory consolidation in senior dogs. In addition, given the marked performance improvement following the 3-hour polysomnography—but not the 1-hour sleep condition—further evaluation is needed to determine the ideal amount of sleep required for adequate memory consolidation.
The effects of social interactions and environment on sleep patterns in dogs are other promising research areas. Dr. Kis mentioned her team’s ongoing research on the effects of positive and negative social interactions on sleep structure in dogs. Regarding an environmental effect, Dr. Kis and her team, in a limited number of observations, have also identified pattern differences in dogs sleeping in a laboratory setting versus in their home environments.
In the real world, understanding the relationship between sleep and learning is applicable to dog training. Allowing time for memory consolidation would help a dog master 1 task before learning another.
Dr. Pendergrass received her DVM from the Virginia-Maryland College of Veterinary Medicine and completed a postdoctoral fellowship at Emory University’s Yerkes National Primate Research Center. Dr. Pendergrass is the founder and owner of JPen Communications, LLC, a medical communications company.
- Vertes RP, Siegel JM. Time for the sleep community to take a critical look at the purported role of sleep in memory processing. Sleep. 2005;28(10):1228-1229; discussion 1230-1233.
- Diekelmann S, Born J. The memory function of sleep. Nat Rev Neurosci. 2010;11(2):114-126. doi: 10.1038/nrn2762.
- Kavanau JL. Vertebrates that never sleep: implications for sleep’s basic function. Brain Res Bull. 1998;46(4):269-279.
- Ákos P, Thalhammer JG, Leschnik M, Halász, P. Electroencephalographic examination of epileptic dogs under propofol restraint. Acta Vet Hung. 2012;60(3):309-324. doi: 10.1556/AVet.2012.026.
- Miklósi A, Topál J. What does it take to become 'best friends'? Evolutionary changes in canine social competence. Trends Cogn Sci. 2013;17(6):287-294. doi: 10.1016/j.tics.2013.04.005.
- Langford FM, Cockram MS. Is sleep in animals affected by prior waking experiences? Anim Welf. 2010;19(3):215-222.
- Gais S, Mölle M, Helms K, Born J. Learning-dependent increases in sleep spindle density. J Neurosci. 2002;22(15):6830-6834.
- McKinney SM, Dang-Vu TT, Buxton OM, Solet JM, Ellenbogen JM. Covert waking brain activity reveals instantaneous sleep depth. PLoS One. 2001;6(3):e17351. doi: 10.1371/journal.pone.0017351.
- Roozendaal B, McEwen BS, Chattarji S. Stress, memory and the amygdala. Nat Rev Neurosci. 2009;10(6):423-433. doi: 10.1038/nrn2651.