Research Methods for Understanding Aircraft Noise Annoyances and Sleep Disturbance (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

PHASE II – AIRCRAFT SLEEP DISTURBANCE STUDY PLANS 14 Sleep Disturbance Literature Review 14.1 Background The principal aim of the literature review was to critically review the scientific literature concerned with the effect of aircraft noise on sleep while additionally explaining sleep assessment concepts for the non- specialist. Particular attention was given to the strengths and weaknesses of the literature with identification of gaps in the present state of knowledge. The ultimate purpose of the review was to provide a background on which to design further studies (described in Sections 15, 16 and 17) which are relevant, robust, cost-effective and achievable. ‘A good night’s sleep’, in terms of adequate duration, depth, composition and continuity is generally accepted as a prerequisite for human health and well-being. Consequently most people would consider a quiet night time for undisturbed sleep as a reasonable expectation from society. The link between aircraft noise and sleep disturbance is clear but the further link between sleep disturbance and long-term health effects, despite mechanisms being suggested, has not been conclusively established. There are some indications that night time noise can be linked to cardiovascular disease and stroke in the elderly but presently there is no clear evidence of a pathway that directly links noise (at commonly experienced levels) and disturbed sleep with cardiovascular disease. The early studies (pre-1990) of noise disturbed sleep tended to be survey reports from residents living near to airports or small laboratory-based reports which showed clear effects of sleep disturbance due to pre-recorded aircraft noise events (ANE) with a direct causal relationship between noise level and the degree of sleep disturbance reported. However, limited subject numbers, lack of adequate controls and no generally accepted data gathering or standardized analytical techniques precluded any clear agreement on the exact relationship between aircraft noise and sleep disturbance in the field. The predominant sleep research technique, over the last 40 years, has been polysomnography, which involves electrophysiological techniques based around electroencephalography (EEG), a measure of the electrical activity of the brain recorded from the scalp by electrodes. This technique is frequently described as the ‘gold standard’ for sleep recording and has given rise to the concept of sleep stages. The EEG provides a clear indication of an individual’s level of physiological arousal. There is a continuum of arousal which ranges from very high levels of alertness to deep sleep. The detailed analysis of sleep involves breaking the continuous electrophysiological recordings into short epochs (typically 30 seconds) and each epoch is ascribed one of seven epoch scores: Wake; Movement Time; stages 1, 2, 3 or 4 sleep, which can be considered as increasing progressively in depth and collectively known as non-REM sleep, while REM (rapid eye movement) sleep is typically associated with dreaming. The combination of sleep stages 3 and 4 can also be described as SWS (slow wave sleep)12. 12 According to a revision of the sleep scoring manual by the American Academy of Sleep Medicine in 2007, SWS would be classified as stage N3. Likewise, Movement Time does not exist in the new classification anymore, and would be classified as Wake instead. 38

In addition to investigating additional evoked arousals due to noise, the assessment of the degree of sleep disturbance involves noting: sleep onset delay, increased awakenings, decreased slow wave sleep13, REM sleep, and early awakenings. Polysomnography is very sensitive, extremely well standardized and the dominant method in modern sleep research particularly for identifying arousals which seem to be essential for the detrimental effects of sleep fragmentation on daytime functioning. However, polysomnography is very time consuming, needs highly specialized equipment and considerable staff expertise for attaching equipment and for data analysis which explains its one major weakness in its considerable expense compared to other simpler but less direct methods of assessing sleep disturbance. This cost tends to limit the sample size which in turn reduces the statistical power and generalizability of the studies, and prevents any firm conclusions being drawn from such studies. In addition, the electrodes attached to the subjects scalp and skin are somewhat invasive and may influence sleep, especially during the first measurement night. Finally, as sleep stages are scored visually by humans, high inter- and intra-observer variabilities have been reported in the past. Therefore, a comprehensive consideration of other techniques available is essential as there is a requirement in this current work to propose alternative research methods for field studies to assess the relationship between aircraft noise and sleep disturbance for US airports. Other less direct methods of assessing sleep disturbance are: 1) Behaviorally confirmed awakening (BCA) The subject ‘pushes a button’ to indicate an awakening, which is inexpensive and has high specificity but its major weakness is low sensitivity, as it will underestimate brief periods of arousal and wakefulness during sleep which may potentially be critical issues in any harmful processes. In addition, this method can be prone to subjective error. 2) Heart rate response (HRR) Involves assessing changes in the autonomic arousal level via the heart rate which on a noise can accelerate and display a heart rate response (HRR). Recent work showed that EEG defined awakenings and cardiac activations were positively related to increasing maximum sound pressure level (SPL) of the ANE and the two different measures of sleep disturbance were highly correlated. The HRR technique has the advantage of being relatively inexpensive, simple and robust particularly when looking for responses to specific events (e.g. aircraft noise). 3) Actigraphy monitors body movements via a movement device, about the size of a man’s wrist-watch attached to a limb (generally the non-dominant wrist.) Inside the device a programmable microprocessor logs displacements of a movement detector. The general assumption is that periods of quiescence are associated with sleep while movement is more likely to be associated with disturbance and wakefulness. This technique is very simple, inexpensive, data rich, has been used in previous large field studies in Europe particularly and has high sensitivity but low specificity similar to HRR. However if these methods are combined the specificity would be improved but both these methods can be criticized for not being able on their own to differentiate between sleep and wakefulness states. 4) Post-sleep questionnaires are common procedures in sleep research and a number of standardized formats are available which are designed in general to assess the quantity and quality of perceived sleep by the subject. The detailed form of the questionnaire is usually tailored to suit the particular objectives of the study. There have been a few large field studies, mainly in Europe, funded by state agencies which have used primarily BCA and actimetry with some polysomnography for calibration, including a major German study which incorporated both laboratory and field studies. These studies showed that awakening probability increases with maximum SPL of the aircraft noise event (ANE). 13 Slow wave sleep - SWS – is generally considered to be the deepest and the most recuperative sleep. 39

There have been many noise-response curves developed which indicate a threshold of significant disturbance where the response moves above background levels of internally generated arousals. The field studies, similar to laboratory studies, indicated a direct causal relationship between noise level and the degree of sleep disturbance. There have been recent attempts to update and enhance the analysis of previously recorded sleep awakening data. However there is still a large gap in our knowledge with an inability to produce a general consensus for a universally accepted noise –response curve that airports and regulatory bodies could use for policy purposes. Previous reviewers on this topic have expressed the difficulties of designing sleep disturbance studies and integrating and summarizing the results of previous studies pointing out:  The different methodologies and analytical procedures have led to a wide range of input and output variables;  Large individual subject differences and predictive relationships that account for only a small fraction of the variance. All of which have led to studies that are seemingly contradictory and inconclusive. However, there is an awareness of a need for further study which would ideally be longer term, employ larger samples and lower cost techniques adequately calibrated against accepted standards. 14.2 Summary of Noise Effects on Sleep In general the following results have emerged from various studies concerned with the effects of noise on sleep:  In a normal quiet night, polysomnography indicates that people experience about twenty arousals that result in brief periods (less than 1 minute) of waking and subjects are usually able to remember about two or three awakenings, when completing a post-sleep questionnaire in the morning.  There is no doubt that awakening probability increases with maximum SPL (Lamax) of the ANE but there are considerable individual differences in the sleep responses.  There are fewer noise related responses in children, but these responses increase with age. There are some indications that males have significantly more discrete movements than women and are more likely to respond to ANEs.  There appears to be an inverse relationship between ambient bedroom noise levels and the chance of awakening due to aircraft noise i.e. in quieter backgrounds subjects are more likely to respond to ANEs. Also, sleep is more likely to be disturbed the longer the quiet period before an ANE.  The size of the motility reaction and awakening probabilities were found to be dependent on the slope of the noise envelope associated with the ANE such that more rapid rise times gave greater likelihood of a motility response. There was more chance of a reaction to noise with a greater amount of high frequency (>3 kHz) components in its spectral composition.  The autonomic responses (HRR) to noise were more sensitive but less specific and did not habituate while cortical (EEG) responses were more specific, displayed dose-response and habituation features. The HRR response is significantly larger if the arousal evokes an awakening response.  Most responses occur in lighter sleep (stages 1 & 2) and REM. Also, awakening probability is lower from SWS (stages 3+4). Noise tends to delay sleep onset and can reduce SWS.  Road traffic causes the most obvious changes in sleep structure and continuity whereas air and rail noise events were considered more disturbing subjectively. Subjective annoyance was greater for aircraft noise, while cortical and cardiac responses (HRR) during sleep were lower for air compared to road and rail traffic. 40

 The cumulative noise energy exposure (e.g. Lnight or Ln) did not predict sleep disturbance, supporting the notion that physiological responses depend on individual sounds not cumulative energy metrics.  The overall correlation between actimetric measures of disturbance (motility) and indoor A- weighted sound exposure level (SEL) of individual noise events has been found to be relatively high, while correlation with measures of behavioral awakening (button pressing) were less.  The filtered actigrams were able to detect 88% of all the EEG-determined periods of interim wakefulness longer than 15 seconds and movement time longer than 10 seconds but there were many false positives.  Actigraphy has shown that below about 80dBA (Lmax) outdoors, there was hardly any increase in actigraphic response above the background movement activity associated with sleep. For ANE above this level there was a response rate of about 1 in 30, i.e. 3.3%.  Only a minority of ANEs affected sleep while domestic and idiosyncratic items had more clear effects. 14.3 Considerations for Study Plan A detailed consideration of recent reviews of this topic revealed considerable agreement on the weaknesses of the field studies to date: too few subjects and too many uncontrolled variables leading to results that tend to be inconclusive and contradictory. Such results are due to the difficulties of controlling such a large number of variables, e.g., individual differences in noise sensitivity; attitudes to airport operations; gender and age. Further complications arose in comparing studies due to different (1) methodologies for recording sleep (2) criteria for defining disturbance in terms of arousals and awakenings (3) analytical procedures employed. The one major gap in this research area is the lack of a relatively inexpensive technique which could provide sufficient data to be able to boost the power of the analysis and offset the possible influence of uncontrolled variables. Most recent reviews on this topic have advocated further field work with larger subject numbers for longer time periods and where possible simpler and more inexpensive techniques appropriately calibrated against traditional methods and building on the research already completed. The literature review identified important associated issues that need careful consideration in planning future studies in this area, including:  Field versus laboratory study, concluding the field (in people’s homes) being the most appropriate location;  Noise recording methodology ideally requires indoor and outdoor (Lamax) and real time sound recording for identification of non-aircraft noises;  Location should provide a range of aircraft noise levels across the nighttime and from subject to subject;  The subject mix should reflect the base population whenever possible with comprehensive data gathered during subject selection. 41