David L. Dickinson

A practical validation study of a commercial accelerometer using good and poor sleepers

We validated a Fitbit sleep tracking device against typical research-use actigraphy across four nights on 38 young adults. Fitbit devices overestimated sleep and were less sensitive to differences compared to the Actiwatch, but nevertheless captured 88 (poor sleepers) to 98u2009percent (good sleepers) of Actiwatch estimated sleep time changes. Bland–Altman analysis shows that the average difference between device measurements can be sizable. We therefore do not recommend the Fitbit device when accurate point estimates are important. However, when qualitative impacts are of interest (e.g. the effect of an intervention), then the Fitbit device should at least correctly identify the effect’s sign.

The viability of an ecologically valid chronic sleep restriction and circadian timing protocol: An examination of sample attrition, compliance, and effectiveness at impacting sleepiness and mood

Chronic sleep restriction (SR) increases sleepiness, negatively impacts mood, and impairs a variety of cognitive performance measures. The vast majority of work establishing these effects are tightly controlled in-lab experimental studies. Examining commonly-experienced levels of SR in naturalistic settings is more difficult and generally involves observational methods, rather than active manipulations of sleep. The same is true for analyzing behavioral and cognitive outcomes at circadian unfavorable times. The current study tested the ability of an at-home protocol to manipulate sleep schedules (i.e., impose SR), as well as create a mismatch between a subject’s circadian preference and time of testing. Viability of the protocol was assessed via completion, compliance with the SR, and success at manipulating sleepiness and mood. An online survey was completed by 3630 individuals to assess initial eligibility, 256 agreed via email response to participate in the 3-week study, 221 showed for the initial in-person session, and 184 completed the protocol (175 with complete data). The protocol consisted of 1 week at-home SR (5-6 hours in bed/night), 1 week wash-out, and 1 week well-rested (WR: 8-9 hours in bed/night). Sleep was monitored with actigraphy, diary, and call-ins. Risk management strategies were implemented for subject safety. At the end of each experimental week, subjects reported sleepiness and mood ratings. Protocol completion was 83%, with lower depression scores, higher anxiety scores, and morning session assignment predicting completion. Compliance with the sleep schedule was also very good. Subjects spent approximately 2 hours less time in bed/night and obtained an average of 1.5 hours less nightly sleep during SR, relative to WR, with 82% of subjects obtaining at least 60 minutes less average nightly sleep. Sleepiness and mood were impacted as expected by SR. These findings show the viability of studying experimental chronic sleep restriction outside the laboratory, assuming appropriate safety precautions are taken, thus allowing investigators to significantly increase ecological validity over strictly controlled in-lab studies.

Personal sleep debt and daytime sleepiness mediate the relationship between sleep and mental health outcomes in young adults

Sleep duration and chronotype (i.e., morningness–eveningness) are associated with increased depression and anxiety risk, but differences in individual sleep need and lifestyle may mean these sleep parameters do not present the same risk across all individuals. This study explored the mediating role of sleep debt and daytime sleepiness in the relationship between sleep and mental health symptoms in young adults, a particularly vulnerable population.

Neural correlates of decision-making during a Bayesian choice task

Many critical decisions require evaluation of accumulated previous information and/or newly acquired evidence. Although neural correlates of belief updating have been investigated, how these neural processes guide decisions involving Bayesian choice is less clear. Here, we used functional MRI to investigate neural activity during a Bayesian choice task involving two sources of information: base rate odds (‘odds’) and sample evidence (‘evidence’). Thirty-seven healthy control individuals performed the Bayesian choice task in which they had to make probability judgements. Average functional MRI activity during the trials where choice was consistent with use of Odds, use of Evidence, and use of Both was compared. Decision-making consistent with odds, evidence and both each strongly activated the bilateral executive network encompassing the bilateral frontal, cingulate, posterior parietal and occipital cortices. The Evidence consistent, compared with Odds consistent, decisions showed greater activity in the bilateral middle and inferior frontal and right lateral occipital cortices. Decisions consistent with the use of Both strategies were associated with increased activity in the bilateral middle frontal and superior frontal cortices. These findings support the conclusion that both overlapping and distinct brain regions within the frontoparietal network underlie the incorporation of different types of information into a Bayesian decision.