Kageyama, Prof T, Expert Review of NHMRC Draft Information Paper

Comment to consultation draft (Feb 2014)

Dr. Kageyama, Takayuki Professor,
Department of Psychiatric & Mental Health Nursing,
Oita University of Nursing & Health Sciences

Email: [email protected]

This draft is produced through a great effort for selecting appropriate literatures and critically summarizing their contents, so that I almost agree with this draft from the viewpoints of study selection, discussion, and conclusion. However, I would like to present some comments to the appraisal, which is mainly written in 3.4, because this draft will be cited by many persons concerned for many purposes, if once published. 

Study design

In general cross-sectional studies have disadvantages to determine causal relationship. However, if an individual has lived near wind turbine generator for a long time (and power level of wind turbine has been stable for a long time), recently measured exposure to wind turbine noise (WTN) level may equal to long-term exposure level. Furthermore, noise does not accumulate in human body. Effects of too much noise on sleep occurs shortly after exposure to the noise. Taking account of the above facts, the weak points of cross-sectional studies on noise and sleep seem to be lessened.

Outcome of sleep

Sleep is assessed in terms of physical/objective and psychological/subjective phenomena. However, we often observe large variance in physiological features of sleep at one night, and the features also depend on measurement condition. This means long-term measurement is required for a study. On the other hand, we can report our long-term state of sleep, suggesting that single question for sleep can be long-term assessment. Therefore the weak points of previous studies, which focused on self-reported measures of sleep quality, are not meaningless.

However, the previous studies have other kinds of weak point. Most of the studies focused on WTN-induced sleep disturbance. However, sleep disturbance is multicausal in general, and it is sometimes difficult for an individual to specify the reason for sleeplessness (Pressman 1996). Questions about WTN-induced annoyance may affect the responses to the following questions about WTN-induced sleeplessness (van den Berg 2012). Furthermore the definition of sleeplessness in the above studies often lacks specifying frequency, or often includes infrequent sleeplessness, e.g. once a month. To define insomnia, researchers should specify frequency and aftereffects of sleeplessness (American Psychiatric Association 1987; American Academy of Sleep Medicine, 2005). In future studies, therefore, sleep disturbance should be assessed apart from WTN, taking account of its frequency and aftereffects. This is more important than using established tools such as PSQI for sleep assessment. 

Confounding factors

Confounding factors are associated with both the exposure and the outcome, being differentiated from modifiers or correlates to the outcome. The systematic review identified a number of modifiers or correlates that might associate with health or health-related outcomes. In these studies, however, the correlations between WTN and these factors were not always reported; namely they were possible confounding factors. If economic benefit from wind farms is not a confounder but a modifier, this factor does not explain the associations between wind farms and health or health-related outcomes. More correct terminology is required.

It also should be noted that a part of the above possible confounding factors are concerning wind turbines. If we want to collect these informations, it is difficult to mask intent of studies. 


I would like to point out a trivial question. Although problems of multiple comparison are discussed in p166, multiple test is more correct than multiple comparison. For example, repeated test for correlation coefficients will raise similar problems to multiple comparison in ANOVA.

Ongoing research

This draft includes introduction to ongoing research (p166-168). I would like to inform additional research in Japan. We have conducted a research project on evaluation of WTN impact on human health over three years from fiscal year 2010, taking the above into consideration, and being funded by the Ministry of the Environment, Japan. This study is partly reported in the following paper;

Tachibana H, Yano H, Sakamoto S, et al. (2013). Nationwide field measurements of wind turbine noise in Japan, Inter-Noise2013.

Kuwano S, Yano T, Kageyama T, et al. (2013). Social survey on community response to wind turbine noise in Japan, Inter-Noise2013.

Yano T, Kuwano S, Kageyama T, et al. (2013) Dose-response relationships for wind turbine noise in Japan, Inter-Noise2013.S.

Results of further analysis will be reported by Kageyama at ICBEN 2014. In this study, sound measurement was performed in field studies, and administered-questionnaires were developed on the basis of previous literatures and international criteria for annoyance, sleep, and self-reported symptoms concerning physical/mental health. Some important confounders, modifiers, or correlates to the above outcomes are taken into account. According to our tentative results, annoyance due to WTN and insomnia regardless of reason for sleeplessness seem to associate with WTN, even after adjustment for main correlates.

If you want to get more information on this study, please contact with Dr. Tachibana, chief in research project.

Download the review as a PDF →