Impact of habitual short sleep and perceived sleep need on risk-taking in young people

Introduction: Acute sleep deprivation has been reported to increase risk taking; however, less known about how the common condition of chronic sleep restriction affects willingness to take risks. Given the increasing prevalence of short sleep duration, we investigated the prospective relationship of habitual short sleep in young people to their performance on a laboratory task that assesses willingness to take risks.
Materials and methods: Participants (n=166, aged 17-25, 63% female) completed a 7-day protocol at two time points, separated by 12 months in which they wore actigraphs and completed a sleep diary for 5 days at Time 1 (T1) and 6 days at Time 2 (T2). On the following day at each time point, they came to the laboratory to complete the Risky Gains Task as an assessment of risk-related decision-making. In this task, participants can make an immediate choice that provides a guaranteed low payout (safe choice) or wait to select one of two higher payout options, which also carry a significant risk of instead producing a loss (risky choice). Sleep measures assessed from actigraphy and diary included average and variability of total sleep time, sleep efficiency and timing of mid-sleep. Based on the National Sleep Foundation's recommended sleep time, those with average sleep durations of < 6 h were classified as short sleepers. We measured participants' perceived sleep need by asking them how many hours of sleep they think they need. Overall number of risky choices and the number of safe choices following punishment were used as outcome variables of risk-taking behaviors.
Results: Perceived sleep need among short sleepers (7.8±1.4 h) was significantly shorter than among normal sleepers (8.3 ±1.5 h; F₁‚₁₆₇=4.043,p=.046). After controlling for demographic factors (age, sex, body mass index,family income) and risk-taking tendency at T1, there was a significantinteraction between habitual short sleep measured by actigraphy and perceivedsleep need at T1 in predicting the number of risky choices at T2 (F₈‚₁₃₉=9.575,adjusted R²=.431, p< .001). There were also significant interactions between bothactigraphymeasured (F₈‚₁₃₉=28.507, adjusted R² =.673, p< .001), andsleep diary-measured (F₈‚₁₃₉=3.974, adjusted R² =.170, p< .001)short sleep at T1 with perceived sleep need in predicting number ofsafe-choices after loss trials at T2. Among short sleepers, higher perceivedsleep need was associated with increased numbers of risky choices and decreasedsafe choices after loss, while normal sleepers showed the inverse relationshipbetween sleep need and risk-taking in that higher perceived sleep needcorresponded to decreased numbers of risky choices.
Conclusions: Our longitudinal data show that the effects of habitual short sleep on risk-taking tendency depend on an individual's perceived sleep need. Short sleepers with greater sleep need seem to be most vulnerable in making risky decisions. Our findings highlight the importance of evaluating individual differences in perception of sleep need in understanding the impact of short sleep duration in both clinical and research settings.
Acknowledgements: This study is funded by the General Research Fund (#17612015), Research Grant Council, HKSAR. Copyright © 2017 World Sleep Society.