Several months ago, Kotaku posted a sleep study conducted by Daniel King (University of Adelaide) and colleagues. The Kotaku piece’s raised some eyebrows in that they referenced to an earlier study conducted by the same lab that “contradicted” the later study. Critical details about both studies were omitted along with the visceral defensive reader comments. What I see here is information filtered through four layers of news. The Kotaku piece heard it from Gamepolitics, which it heard from medicalxpress.com who reported from a press release by Flinders University. At every stage, some information is lost and new ones are created from reporters’ interpretation. Well, I am going straight to the source and review it.
Video-gaming is an increasingly prevalent activity among children and adolescents that is known to influence several areas of emotional, cognitive and behavioural functioning. Currently there is insufficient experimental evidence about how extended video-game play may affect adolescents’ sleep. The aim of this study was to investigate the short-term impact of adolescents’ prolonged exposure to violent video-gaming on sleep. Seventeen male adolescents (mean age = 16 ± 1 years) with no current sleep difficulties played a novel, fast-paced, violent video-game (50 or 150 min) before their usual bedtime on two different testing nights in a sleep laboratory. Objective (polysomnography-measured sleep and heart rate) and subjective (single-night sleep diary) measures were obtained to assess the arousing effects of prolonged gaming. Compared with regular gaming, prolonged gaming produced decreases in objective sleep efficiency (by 7 ± 2%, falling below 85%) and total sleep time (by 27 ± 12 min) that was contributed by a near-moderate reduction in rapid eye movement sleep (Cohen’s d = 0.48). Subjective sleep-onset latency significantly increased by 17 ± 8 min, and there was a moderate reduction in self-reported sleep quality after prolonged gaming (Cohen’s d = 0.53). Heart rate did not differ significantly between video-gaming conditions during pre-sleep game-play or the sleep-onset phase. Results provide evidence that prolonged video-gaming may cause clinically significant disruption to adolescent sleep, even when sleep after video-gaming is initiated at normal bedtime. However, physiological arousal may not necessarily be the mechanism by which technology use affects sleep.
I started following a faculty member’s advice of avoiding screen media in the morning as it was distracting and my productivity improved! Maybe if we can compare people who play first thing in the morning vs. those who don’t…
The literature review is quite short as there is scant research on videogames’ effect on sleep. Some studies were surveys and there are five experiments including the one I’m reviewing. Some explanations on the adverse impact of videogames on sleep include displacement of normal sleep time, increased mental, emotional or physiological arousal and bright light exposure causing delay of circadian rhythm, that is our biological chronometer is affected by artificial lighting, it is also responsible to tell the body when to wake up based on the rise and fall of the sun. However, there is little to empirical findings that explain why and how media affects sleep.
The four experiments involved relatively small number of participants, about a dozen or less. I contacted the lead author, Dr. Daniel King, for the reasons of their study’s small sample. Quite simply, the cost of running a sleep experiment is expensive: compensation for the participants, having them stay over at the sleep lab for two nights, the equipment costs, the personnel to run the experiment (3-person team) and to analyze the data. Furthermore, other experiments ran with fewer participants, so a sample size criticism to their study is also a criticism to the research area. And this is why science needs money to properly run studies and don’t tell us to run with less, you don’t know what’s it like.
The authors focused on the studies’ findings of sleep-onset latency, that is how long it takes from being wakeful to being sleepy. The findings indicate that playing videogames add a few minutes longer to people’s sleep-onset latency, from 2.3 minutes (Higuchi et al., 2005) to 22 minutes (Dworak et al., 2007). One study found that playing videogames has no effect on sleep with the exception of going to bed later.
The authors observed that playing videogames for 50-60 minutes may be insufficient to produce noticeable effects. So, they pushed to see where it starts to affect sleep by comparing participants playing 50 minutes versus 150 minutes of videogames. They hypothesized that prolonged (i.e., 150 minutes) of videogame play would increase sleep-onset latency and reduced total sleep time which is explained by heightened physiological arousal.
Participants: 17 male high school students who regularly play videogames for 17.7 hours per week. Average age is 16. Other details include that they suffer from any sleep disorder, are in good health, not taking medications, etc.
Videogame used: Warhammer 40,000: Space Marines played on the Playstation 3. The videogame was picked because at the time of experiment, it was just released, so none of the participants had any prior experiences with it. The authors detailed the conditions of the room, such as temperature, brightness and sound level.
The procedure is very detailed and is very controlled. I recommend that you read the article itself for the details and the polysomnography wikipedia article for background information. I will only describe the testing night procedure.
Participants come in to the sleep lab for two testing nights, separated by a week. Participants come to the sleep lab from school at 3:30pm. Their physical activity, naps and caffeine consumption restricted. At 6pm, they eat dinner and changed into night gear where heart rate monitors and sleep monitors is affixed to them. Each participant has their own private bedrooms.
For one testing night, participants either played 50 or 150 minutes before their usual bedtime as determined by their sleep diary before the start of the experiment. On their second testing night, they played the other length of time. So, if they 50 minutes on their first night, then they played 150 minutes on their second night. In a sense, the participants served as their own comparison group. When participants wake up, they completed a set of questionnaires, one particular question is how longer they want to play the videogame.
The authors detailed their data recording and analysis that should interest any sleep researchers. The only thing I understood is, in addition to the sleep and heart rate monitors, they also had participants fill out questionnaires about their subjective level of sleepiness, mood after videogame play among other things.
Their analyses from participants’ polysomnograms revealed numerous non-significances and two significant differences. Sleep onset latency, stage 1 to REM sleep stage were non-significant. All participants took about 15 minutes to fall asleep. Participants in the 50 min condition (M = 12.6, SD = 13.3) vs. 150 min. (M = 16.1, SD = 19.4). The total sleep time was significant different when participants played for 150 minutes, they slept for 391.3 minutes (SD = 49.1) compared to when they played for 50 minutes, they slept for 418.6 minutes (SD = 42.9). There is a significant difference in sleep efficiency, sleep efficiency is 88.9% (SD = 8.8) for the 50 minutes game time, but efficiency fell to 81.9% (SD = 12.9) for the 150 minutes game time. A cause for concern where 85% or higher is considered normal sleep efficiency.
Participants subjective evaluation of their sleep experience revealed only one significant difference. When they played for 150 minutes, they felt it took 39.1 minutes (SD = 36.1) to fall asleep compared when they played for 50 minutes, they felt it took 22 minutes (SD = 18.8) to fall asleep.
Heart rate differences between 50 minutes and 150 minutes of videogame play were nonsignificant, even all participants slept at their usual bedtime. Participants’ moods between conditions were not different, except that when participants played for 50 minutes, they felt they have not played long enough and would like to play an extra 37.7 minutes (SD = 43.4) compared to when they played for 150 minutes, when they would like to play an extra 14.1 minutes (SD = 18.9).
The take home message is that 150 minutes of videogame play caused a clinically significant reduction of sleep time for adolescents. Playtime of 50 minutes affects sleep well within acceptable range. If you want to enjoy a videogame and a good night sleep, best to play one hour-long match or two half-hour matches before your usual bedtime.
The results on their physiological arousal data showed that the participants were playing in a resting state, so they are pleasantly and calmly enjoying the game. The authors argued that physiological arousal could not explain the differences they found and suggest other mechanisms at work. However, we should not throw out physiological arousal as a causal factor, perhaps as a causal factor for a certain range of situations. Despite the videogame being fast-paced and violent, how it is played might affect physiological arousal or should I say with whom. Perhaps some play time in League of Legends might put the participants in a state of high arousal, especially if the researchers recruited participants who regularly play it. Let’s see how winning vs. losing affect their sleep.
The authors discussed some of the consequences of prolonged (i.e., 150 minutes) of videogame play. A consistent pattern of playing long hours may result in poor sleep quality and chronic sleep reduction which can affect academic performance. Interestingly, their data tells us about our ability to monitor how much time we play, which seems poor to me. Participants during the 50 minutes condition wanted to play, on average, 37 minutes longer compared to the 14 minutes in the 150 minutes condition. Were these participants played unmonitored, the might have played well past their bedtime which affect their sleep quality. I probably say some videogames would have some quality that makes us say “one more turn”, “5 more minutes” or “I’ll stop until I finished this quest” which keeps us awake until we have closure.
However, I might add a qualification in that these participants played a just-released videogame and this might due to a novelty effect. I have heard stories of gamers binge playing (i.e., playing on for whole days) on new releases for a few days before returning to their normal schedules. It is possible that these participants are under a binge state and would want to play it into completion. I experienced such binge gaming through my first sittings of Mass Effect, Hellgate: London and Space Pirates and Zombies (on Thanksgiving break) before the novelty wears off. A suggestion for another study would be to try a videogame released 6 months ago and familiar to the participants.
The authors discussed the study’s limitations. The obvious is the experimental conditions that the participants are under since they slept at the lab with medical measurements affixed to their bodies, their environmental conditions controlled and participants knowing that they are playing for research purposes. There was a no-videogame control condition, but then… can’t we just compare it with their previous study (Weaver et al., 2010)? Their sample size is small, but then so are the previous studies and given the costs to conduct a sleep study. The lack of generalization beyond male adolescents with no history of sleep difficulties and regularly plays videogames. The authors discussed that videogames would have smaller effects among older populations, but have greater effects among younger populations.
I’d like to compare between the Kotaku piece with the article and see the differences in reporting. Kotaku did not explicitly reported how long prolonged gaming is, just leaving a link to Gamepolitics who reported the amount of time. An attentive reader could figure it out through Dr. Gradisar’s quotes. Nevertheless, my reading of the piece was framed as if the studies were contradicting each other. Kotaku correctly reported the loss of sleep time of 27 minutes, however they misreported the amount of time of how long participants took to fall asleep. The article stated participants’ self-report of falling asleep on average of 39 minutes, which is different from the objective measure of 16 minutes. The Kotaku piece made no such differentiation whether it was self-reported or not.
Flinders University’s press release and Kotaku mentioned the difference of REM sleep, which is that 150 minutes caused a loss of 12 minutes of REM sleep. According to the article, it’s a non-significant difference. The discussion section made a few notes about REM difference having a moderate effect and is similar to previous studies, but the statistical result was non-significant. The press release and Kotaku did not explain how such reduction of REM sleep had any effects, merely giving background information about REM sleep’s role for cognitive performance. Such piece of information could lead readers to interpret it having a causal and negative effect. Finally, there were no reports about the study’s physiological results from Kotaku or the press release. That is the result of information passing through multiple reporting agents.
King, D. L., Gradisar, M., Drummond, A., Lovato, N., Wessel, J., Micic, G., Douglas, P., & Delfabbro, P. (in press). The impact of prolonged violent video-gaming on adolescent sleep: an experimental study. Journal of Sleep Research, (p. n/a). DOI: 10.1111/j.1365-2869.2012.01060.x