The effects of noise disruption on sleep disturbance

These experiments are specifically designed to establish wind farm versus traffic noise effects on sleep efficiency (macro-structure), and micro-arousal responses to each noise type at equivalent A-weighted Sound Pressure Levels (SPL). Twenty participants from each of the 4 experimental groups will be asked to attend the laboratory for 7 consecutive night sleep studies (4 experimental nights, each with an intervening recovery night). Experimental nights will consist of; prolonged wind farm and traffic noise exposures (WF and T nights), a noise fragmentation night (i.e., randomised and counterbalanced WFN and TN at various SPL, along with silent control exposures), and a control night in random order. Participants will be instrumented for sleep and physiological measures prior to going to bed at their habitual bedtime each night and instructed that they may or may not notice any noises on any of the nights. Experimental noise exposures will only commence once polysomnographically defined stable (>2 min) stage 2 or deeper sleep is already established (or re-established) to avoid noise interference with sleep onset, which we do not believe can be adequately blinded in full night noise exposure experiments. WFN and TN arousal thresholds in deep and stage 2 sleep will be measured on each experimental night using an established brief noise-escalation protocol. WFN and TN nights will be constructed to approximate worst case noise exposure nights (from existing recordings) using 3-min periods of pre-recorded noises with minimal time-variability, randomised into a sequence containing both no noise (control exposures) and A-weighted SPLs up to original recording maxima in 3-6 dBA increments. We will use fade-in over the first 30-sec and A-weight SPL matching between noise types and include noises both unfiltered and filtered to exclude low frequencies <160 Hz and infrasound <20 Hz. The primary outcome is sleep efficiency (sleep time as a % of each 5 min exposure), which requires fixed/ongoing exposures irrespective of arousals or awakenings. On sleep fragmentation nights pre-recorded short (~20-sec) wind farm and traffic noise events will be replayed to more specifically examine physiological activation responses to brief noise events. Noises will be replayed with varying SPLs and filtering similar to WFN and TN nights to span the full range of physiological activation responses from no discernible to “sub-cortical”, micro-arousal and full-awakening responses, with as many replicates as possible throughout the night. This protocol is specifically designed to allow construction of noise dose (SPL)-response curves as a function of stimulus type and sleep stage. In the morning, participants will be asked how well they slept compared to their usual sleep, and to recall what type and number of sounds if any they heard over the previous night to test blinding effectiveness. Salivary samples for cortisol measurements will be collected over the first hour and at 12 hours after initial morning awakening. Between salivary samples in the first hour participants will also undergo a brief psychomotor vigilance task (PVT), and measures of daytime sleepiness. Daytime in-laboratory noise detection and perception tests will commence following breakfast, after which participants will be free to leave the laboratory for the remainder of the day before returning for their next study night. Daytime in-laboratory noise detection and perception tests will compare perceptual responses (acuity, annoyance, loudness and level of acceptance for sleep) and measures of physiological disturbances (heart rate, blood pressure, vasoconstriction responses) of a range of pre-recorded wind farm and traffic noise samples replayed in random order to participants from each survey respondent group. Methods are based largely on previous studies assessing noise annoyance, with the addition of EEG and cardiovascular measurements.