Health effects of wind turbines: a review of the literature between 2010-2020

ABSTRACT

Although wind power is more acceptable in terms of its environmental impact, possible risks to human health are still being discussed. The aim of this study is to systematically evaluate the methodology and the outcomes of the articles that investigate the health effects of wind turbines on humans. Combinations of keywords were entered into the PubMed database. The search resulted in a total of 141 hits, 22 were included. It had been noticed that the most common problems in those living around the wind turbines are noise annoyance(n=18), risk perception and attitude towards wind turbines(n=11), general health symptoms and quality of life(n=11), sleep disturbance(n=10), annoyance(n=7) and shadow flicker effect(n=4). General annoyance is adversely affected by the noise level and sensitivity to noise. We can conclude that the knowledge of and attitude towards wind turbines can turn into annoyance and symptoms if the audio-visual effects of turbines limit daily life activities.

KEYWORDS:

• Health effects
• noise annoyance
• sleep disturbance
• wind turbines

Introduction

Energy is one of the basic indicators reflecting the economic and social development potential of the country, which is an essential element of daily life. One of the consequences of the oil crisis in the 1970s in the context of energy security was that it created a quest for alternative sources such as wind energy (Kılıç et al. 2017). Environmental concerns are another reason for choosing renewable sources of energy such as wind, solar and biomass. Among renewables, wind energy has the second largest capacity after hydro energy (International Energy Agency 2018). Wind power plants (WPPs) do not cause air, soil and water pollution and have the lowest carbon emission compared to other renewable sources (Guerrero-Lemus and Martínez-Duart 2013). Since the wind source is readily available in nature at all times, producing electricity from wind has zero source cost. Although WPPs are more acceptable in terms of their environmental impact, possible risks to human health are still being discussed.

Michaud indicated that knowledge of and attitude towards wind energy can turn into displeasure and annoyance if the audio-visual effects of turbines limit daily life activities (Michaud et al. 2016c). The majority of studies on the health effects of wind turbines (WT) show that personal perspectives, psychological factors and general attitude about turbines are effective on the occurrence of negative risk perception (Pedersen and Waye 2004; Paller 2014). Regardless of the actual noise, general perception about WT can be affected by confounders such as noise sensitivity, perspective about WT, economic benefits, existing health complaints, visual concerns and environmental expectations (Pedersen and Waye 2007; The council of Canadian Academies 2014). Depending on their ability to cope with the physiological or psychological responses to this process, individuals either achieve harmony with the environment or start to develop symptoms related to environmental stress (Michaud et al. 2016c).

Recently, the scope and the content of the research on WPPs and health have been expanding. So, the amount of research has risen with the increasing number of WT installations and their growing proximity to residential areas (Knopper and Ollson 2011; Schmidt and Klokker 2014; Onakpoya et al. 2015; McCunney et al. 2014; Basner and McGuire 2017). Noise and additional problems it causes are the most frequently reported complaints by those living around the turbines (Michaud et al. 2016c). Chronic noise interrupts daily individual activities such as communication, attention, adaptation, sleep, and relaxation; it can also lead to anger, burnout, displeasure, and stress-related symptoms. Features such as the continuity of WT noise (WTN) (day and night), the visibility of the source and the absence of masking background noise (generally installed in semi-rural areas away from the noisy city centre) cause disturbance (Basner et al. 2014). Stress hormone (catecholamine and cortisol) levels increase as hormonal response to noise (WHO Regional Office for Europe 2009). In addition, with the activation of the autonomic nervous system, the heart rate increases and blood pressure rises as a result of peripheral vasoconstriction. According to Pedersen, the risk perception and visual aspects of turbines constitute an important confounding factor for noise disturbance, significantly increasing perceived noise levels and causing deterioration in sleep quality, negative mood, and stress reporting (Chapman et al. 2013; Maffei et al. 2013). Adverse health effects are observed above 40 dBL_{night,outside}, including self-reported sleep disturbance, environmental insomnia, and increased use of sleeping pills and sedatives. Therefore, the World Health Organization recommends the level of 40 dBL_{night, outside} as being equivalent to the lowest observed adverse effect level (LOAEL) for night noise (WHO Regional Office for Europe 2009). Chronic WTN can directly affect sleep, and it can also cause sleep disturbance because of feeling of annoyance (The council of Canadian Academies 2014). One of the visual effects of turbines is shadow flicker and light reflection. Rotating blades of RTs can cause shadow flickering and shadow vibration at sunrise and sunset (Senel and Erdem 2016). If these vibrations fall on the residences, they may cause annoyance in individuals.

This review aims to systematically evaluate the main methodological characteristics of studies published in the last decade that investigate the effects of WTs on the health of people living near a WT. In addition, it investigated the outcomes of the reviewed articles on the following topics: noise annoyance, risk perception, health symptoms and quality of life, sleep, annoyance, and flicker effect.

Material and methods

Study selection criteria for this review

In accordance with the purpose of this review, the studies including WTs and all possible health-related outcomes (noise annoyance, risk perception and attitude toward WT, general health symptoms and quality of life, sleep disturbance, annoyance, shadow flicker effect) were chosen as selection criteria. Study design types (cross-sectional, cohort, case-control, ecological, experimental and qualitative studies) and the language (English) were the other selection criteria. Study designs other than original articles such as editorial letters, reviews and case reports were excluded. Articles published between 2010 and 2020 were considered for inclusion.

Search methods for identification of studies

A literature search was conducted by two researchers for peer-reviewed articles published in English up to September 2020. The terms (‘wind turbine’ OR ‘wind energy’) AND (‘health effects’) were entered into the database PubMed. Articles had to meet relevance criteria for study design, language, year and outcomes. The search resulted in a total of 141 hits. Information regarding any evidence of health-related effects associated with WT was extracted from the reviewed articles by one reviewer and confirmed by the second reviewer. Disagreements were resolved through consensus.

Data collection

The titles and abstracts of these articles were reviewed, 7 duplicates and 76 irrelevant (55 irrelevant, 9 wildlife effects, 3 editorial letters, 3 occupational health risks at WPP, 3 Japanese, 1 Russian, 2 unavailable full texts) articles were excluded. Among the remaining studies, 58 were associated with the topic of adverse health outcomes of WTs. At this stage, reviews (n = 25), guidelines (n = 5), case reports (n = 2), methodological studies (n = 2), newspaper scans (n = 1), and a study that was only about specific outcomes of cochlear implant patients were not covered in this review. At last, 22 peer reviewed articles were included. The flow chart of the selection is shown in Figure 1. All of the selected articles were classified into main topics according to health effect outcomes and then they were examined in detail. Articles on the same study population were included in the review to present additional evidence for health-related effects related to WTN if this evidence was not previously reported in other publications. This review was conducted according to the evaluation process outlined in the PRISMA checklist (Page et al. 2021). The methodological preparation stage including the determination of how to examine the articles and the definition of the main characteristics was based on PRISMA principles.

Figure 1. The Flow Diagram of the Reviewed Articles (PRISMA Flow Diagram).

Results

Technical characteristics of the reviewed articles

The variables of the articles under review were evaluated with different methods and measurement tools. The different characteristics of the articles reviewed herein are presented in Table 1. Accordingly, six studies were conducted in Canada. Eleven were cross-sectional, which is the most frequent study method. Face-to-face interviews and mail surveys were the most commonly used data collection methods. It has been noted that the noise data used for most of the studies was obtained either from records or through measurements. Conflict of interest was stated in only two studies, whereas no information was provided on the topic in three studies. Seven studies did not mention limitations. Public (government or municipality) support was reported in 7 studies and 4 studies highlighted support by the private sector.

Table 1. Main characteristics of reviewed articles.

No	Author	Country/ Year	Study Design	Number of Participants/ Response Rate	Data Collection Strategy	Variables and Tools used to measure	Evidence of Competing Interest	Evidence of Bias/ Limitation	Funding of the Research
1	Jalali et al.	Canada/ 2016a	Cohort	16/ Not indicated	Not clear	Noise: Measurement Sleep: EEG^a, PSG^b, diaries noise sensitivity, attitude to WTs	No	Potential limitation from lack of control group	Municipality and industry
2	Jalali et al.	Canada/ 2016b	Cohort	64/ 30%	Mailbox and door-to-door	Noise: Measurement Sleep: PSQI^c, ESS^d, ISI^e, economic benefit	No	Recall and non-response bias	Industry
3	Michaud et al.	Canada/ 2016a	Cross-sectional	1238/ 78.9%	Interview (face-to-face and telephone)	Noise: Measurement Noise annoyance Sleep: PSQI, Actigraphy, Perceived stress, Lifestyle behaviours, Chronic disease	No	Confounding bias lack of appropriate study design	Public
4	Michaud et al.	Canada/ 2016b	Cross-sectional	1238/ 78.9%	Interview (face-to-face at home)	Noise: measurement Sleep: PSQI Annoyance	No	Sample size homogeneity distribution and noise exposure problems	No
5	Kageyama et al.	Japan/ 2016	Cross-sectional	1079/ Participants 49%, Control 45%	Interview (mail and face-to-face)	Noise: Measurement, Noise sensitivity Sleep: Questionnaire Physical/mental health Environmental Problems	No	Confounding bias	Public
6	Nissenbaum et al.	USA/ 2012	Cross-sectional	79/ Not indicated	Not clear	Noise: Recordings (obtained from publicly available sources) Sleep: PSQI, ESS, mental and physical health: SF-36v2 Attitude toward IWTs, psychiatric disorders	No	Reposting and selection bias	Not indicated
7	Shepherd et al.	NewZealand/ 2011	Experimental	306/ 32–34%	Self-administered (postal)	QOL: WHO-QOL BREF ^f, Annoyance: NPS^g, Environmental amenity, Noise sensitivity	No	Lack of sample size of turbine group and coincident noise measurements were not obtained	Not indicated
8	Mroczek et al.	Poland/ 2015	Cross-sectional	1277/ 85%	Interview (door to door)	QOL^h: SF-36v2 Smoking: Fagerstrom questionnaire Health problems, Risky drinking, Reaction to WT	No	Not indicated	Not indicated
9	Pawlaczyk-Łuszczyńska et al.	Poland/ 2014a	Cross-sectional	156/ 71%	91.7% self-administered (mail) and 8.3% interviewed (some elderly people)	Attitude towards WT, Environmental satisfaction, Chronic illnesses, General wellbeing, Sleep Quality Mental health status: GHQ-12^i Noise: measurement	No	Not indicated	Public
10	Pedersen et al.	Holland/ 2009	Cross-sectional	725/ 33%	Postal questionnaire	Response to wind turbine noise, Economical Benefit Noise: measurement	Not indicated	Not indicated	European union and industry
11	Maffei et al.	Italy/ 2013	Experimental	34/ No sampling	Self-reported	Noise: 3D Virtual Scenarios Visual and acoustic aspects	No	Not indicated	Not indicated
12	Klæboe et al.	Norway/ 2016	Cross-sectional	90/ 38%	Interview (web-based tool)	Living environment, health, annoyances, noise sensitivity Noise: measurement	No	Not indicated	None
13	Magari et al.	USA/ 2014	Cross-sectional	62/ No sampling	Interview (door to door)	response to WT, annoyance, Noise: measurement	No	Not indicated	Not indicated
14	Crichton et al.	New Zealand/ 2014a	Experimental	54/ No sampling	Self-reported	Physical symptoms, blood pressure and heart rate measurement	Not indicated	Not indicated	Not indicated
15	Blanes -Vidal et al.	Denmark/ 2016	Cross-sectional	454/ 40.5%	Self-administered (mail)	Concerning symptoms, perceived environment, personal characteristics, physical symptoms, annoyance, distance to closest WT	No	Confounding selection and non-response bias	Not indicated
16	Crichton et al.	New Zealand/ 2014b	Experimental	60/ No sampling	Self-reported	Symptoms and mood	No	A limitation for noise exposure simulation was indicated	Not indicated
17	Clark et al.	Australia/ 2016	Qualitative	16/ No random sampling (snowballing approach)	Face-to-face in-depth interview	Health complaints	No	Limitations about sampling	Public
18	Walker et al.	Canada/ 2011	Qualitative	178/ No sampling	Face-to-face in-depth interview	Physical Health Complaints about WT	Not indicated	Selection bias	University
19	Chapman et al.	Australia/ 2013	Ecological	129/ 30%	Recorded data	Complaints about noise or health	No	Study design limitations	None
20	Smith et al.	Sweden/ 2020	Experimental	50/ No sampling	Postal, mailings, telephone contact, and public advertising	Sleep: EEG ^j, PSG, EOG ^k Salivary cortisol measurement. Tiredness, tension, irritation, sleep quality, Sleep disturbance by WTN	Yes (one of the author has clients from WT industry)	Self -selection bias, limitations about confounding factors and non-comparability bias	Public
21	Barry et al.	Canada/ 2018	Cross-sectional	1,238 / 78.9%	Secondary data	Distance QOL: WHO-QOL	Yes (three out two authors)	Study design limitations, volunteer bias, ‘survivor bias’	Industry and public
22	Poulsen et al.	Denmark/ 2019	Cohort	711,249/Sample population was determined by noise levels.	Register-based data	Noise: modelling MI ^l or Stroke Cases: ICD 10 code records	No	Limitations about confounding, potential bias from missing data	Public

^a:EEG (electroencephalogram), ^b:PSG (polysomnography), ^c: PSQI: Pittsburgh Sleep Quality Index, ^d: ESS (Epworth daytime Sleepiness Scale), ^e: ISI (Insomnia Severity Index), ^f:WHO-QOL BREF (World Health Organization- Quality of Life short version), ^g:NPS (Neighbourhood Problem Scale), ^h:QOL(Quality of Life), ⁱ:GHQ-12 (Goldberg General Health Questionnaire), ^j:EKG (electrocardiography), ^k:EOG (electro-oculogram), ^l:MI (myocardial infarction)

Summary outcomes of reviewed articles according to the main topics

The summary of the outcomes of the studies is presented in Table 2.

Table 2. Summary outcomes of the reviewed studies.

No	Author	Summary of Outcomes
1	Jalali et al. (a)	There was no significant difference in noise measurements before and after installation, but a significant decrease was detected in sleep quality.
2	Jalali et al. (b)	Negative attitude, visual effects, and concern about property value were associated with poor sleep quality.
3	Michaud et al. (a)	Outdoor WTN up to 46 dB(A) was not associated with subjective sleep disturbance or objectively measured sleep outputs such as sleep latency, sleep efficiency, total sleep time, rate of awakening bouts, and wake duration after sleep.
4	Michaud et al. (b)	Annoyance due to all investigated effects about WTs (noise, vibration, shadow, warning lights, visual effect) increased as the WTN increased.
5	Kageyama et al.	Those who were concerned about environmental issues and expressed auditory and visual annoyance related to WT reported more complaints about insomnia and symptoms. Regardless of the actual noise level, noise sensitivity and visual annoyance were associated with poor health.
6	Nissenbaum et al.	WTN can cause sleep disturbances and mental problems at a distance of 1.4 km.
7	Shepherd et al.	General health of the case group living within 2 km of a turbine was not affected but they did report lower quality of life. Sleep quality deteriorated in correlation with dose and distance.
8	Mroczek et al.	Residents of places where wind-farm developments were either at planning stage or under construction had the lowest overall QoL and general health scores. It has been explained by environmental stress and its adverse health effects.
9	Pawlaczyk-Łuszczyńska et al.	Subjective factors such as noise, general attitude towards WTs and sensitivity to landscape littering might explain 63% of the WT-related annoyance.
10	Pedersen et al.	WTN was found to be more annoying than the other noise sources like transportation and industry because of its high visibility, temporal variability, and lack of night-time abatement characteristics.
11	Maffei et al.	Distance is found to be the strongest factor in determining personal responses. More positive attitude towards green and white turbines has been reported.
12	Klæboe et al.	WTs have been defined as visually poor and non-aesthetic. Non-acoustic factors play an essential role in general perception against WT.
13	Magari et al.	Noise levels have not been associated with annoyance or satisfaction with the living environment.
14	Crichton et al.(a)	Concerns about the health effects of WTN could be explained by psychological expectations.
15	Blanes -Vidal et al.	When checked against the confounders (environmental sensitivity, personal reactions etc.), any significant association between WT and health symptoms disappeared.
16	Crichton et al. (b)	Individual perspectives, negative information and attitudes towards WT could determine physical symptoms.
17	Clark et al.	Participants’ stake, interest and legitimacy characteristics may be relevant with reported WT related health effects.
18	Walker et al.	No significant associations between WT and daily life were observed. During the planning process, a healthy community relationship and appropriate information techniques reduced the psychological outcomes
19	Chapman et al.	The prevalence of the complainants residing around wind farms is 0.4%. When the historical and geographical variations in complaints are investigated, it is seen that nocebo effects play an important role in the aetiology of complaints.
20	Smith et al.	WTN have caused longer rapid eye movement (REM) sleep latency and lower amount of REM sleep. Self-reported sleep was consistently rated as worse following WTN exposure nights.
21	Barry et al.	Residential proximity to WT is correlated with annoyance and health-related quality of life measures.
22	Poulsen et al.	No evidence was found of any association between WTN and MI or stroke.

The articles included in the review were classified as per the main topics, which are WT-related health outcomes. The main topics comprised noise annoyance (n = 18), risk perception and attitude towards WT (n = 11), general health symptoms and quality of life (n = 11), sleep disturbance (n = 10), annoyance (n = 9), and shadow flicker effect (n = 4).

Noise and noise annoyance

Eighteen studies have mentioned about the effects of wind turbine-related noise. These adverse effects are mostly about noise annoyance.

Objective sound measurements were carried out in 7 studies. In some of these studies sound levels were recorded inside the households and even at bedside (Jalali et al. 2016a) while in some others, short-term recordings were performed at randomly selected points (Pedersen et al. 2009; Magari et al. 2014; Pawlaczyk-łuszczyńska et al. 2014; Kageyama et al. 2016; Klæboe and Sundfør 2016; Michaud et al. 2016b). An examination of all studies with reported sound levels revealed that noise levels in the vicinity of WT ranged between 31 and 48 dB, approximately.

Pawlaczyk-ŁUszczyńska’s (Pawlaczyk-łuszczyńska et al. 2014) findings show that 60.3% of the participants were aware of the outdoor WTN and 33.3% of all reported annoyance about it. 39.7% of the participants were also aware of the indoor WTN and 20.5% of all reported annoyance about it. Although the annoyance and awareness of noise decreased both indoors and outdoors with increasing distance, no significant difference was found. 45–50 dB of WTN was the most noticeable and annoying range both indoors and outdoors. This study found WTN to be more disturbing than other environmental noise sources. Magari (Magari et al. 2014) reported no significant difference between noise levels and annoyance. In a study by Michaud et al. (Michaud et al. 2016b), noise annoyance among the participants increased significantly starting from 35 dB. WTN annoyance was reported mostly during summer months, evenings, and nights. Results of the Blanes-Vidal study presented that (Blanes-Vidal and Schwartz 2016) 27% of the participants stated that they were disturbed by the noise outdoors while only 7% of them attributed their annoyance to the WTs. Jalali et al. (Jalali, Nezhad ahmadi, et al., 2016) have found no significant change in indoor and outdoor noise measurements before and after installation. In the study of Nissenbaum, health impacts due to WTN have been hypothesized, but no evidence exists to support a connection except for self-reported sleep disturbance (Nissenbaum et al. 2012). Pedersen’s findings revealed that annoyance with WTN is related to a negative attitude towards the source and to noise sensitivity. Also, it is remarkable that economic benefit, which substantially decreases the probability of annoyance, has the highest impact on noise annoyance between the non-acoustical factors (Pedersen et al. 2009). In the study of Smith et al., the comparison of the control and the WTN-nights after the WTN-nights participants reported more tiredness, irritation and lower pleasantness (Smith Michael et al. 2020).

The noise related WT was the main variable in the rest of the studies and the relationship between WTN, and the different health outcomes were examined. They had been evaluated in the other main topics (Shepherd et al. 2011; Maffei et al. 2013; Crichton et al. 2014a, 2014b; Kageyama et al. 2016; Klæboe and Sundfør 2016; Michaud et al. Michaud, et al., 2016a; Jalali et al. 2016b; Barry et al. 2018; Poulsen et al. 2019).

Risk perception and attitude towards WT

Eleven studies reported data on the general risk perception and attitude of those living near a wind farm. Risk perception is mostly occurring due to the visual impact of WT on the living environment, self-reported sensitivity to noise, personal characteristics, or environmental concerns.

In an experiment by Crichton et al.(Crichton et al. 2014b) participants were divided into two groups so as to create positive and negative expectations. After both groups were exposed to WTN (43 dB) and infrasound noise (9 Hz, 50.4 dB), 24 physical and 24 mental symptoms were investigated. As a result, a significant increase was noted in symptom reporting and intensity among the negative expectation group during the pre- and post-evaluation, while the positive expectation group reported significant decrease in the opposite direction. This study has shown that the information about WT shared on TV shows or the Internet, which are easily accessible by people, are influential in creating and reporting health symptoms. In the study conducted by Maffei (Maffei et al. 2013), WT sites with different features in terms of distance, number and colour were created in a three-dimensional virtual reality environment, and the participants were asked to make audio-visual evaluations. Distance was the most effective factor determining reactions regarding WT. It has been reported that as the distance increased (150 m-500 m), general attitude and visual satisfaction improved; perceived noise, noise disturbance, and noise-induced stress decreased. Green turbines were considered more acceptable and compatible with nature, compared to the brown and red ones. In the study of Chapman et al. (Chapman et al. 2013), complaints, court records, investigations, and news articles about 1634 turbines in 51 WT sites operating between 1993 and 2012 were examined from a psychological viewpoint. 33 out of 51 sites (64.7%) did not receive any complaints. It was observed that a total of 129 (0.4%) complaints were registered from 32,789 people, of which 94 (73%) came from people living around the same 6 sites and associated with anti-turbine groups. As a result, researchers defined these complaints as ‘contagious’ and attributed them to psychological factors. In the study conducted by Pedersen et al. (Pedersen et al. 2009) 14% of the participants reported negative general attitude and 36% reported negative visual attitude. General attitude towards turbines was associated with annoyance rather than noise levels. The strongest correlation was found between visual attitude and annoyance. Magari et al. (Magari et al. 2014) reported that 44% of the participants expressed a positive opinion about WT. The majority (73%) were not aware of the WT or were not disturbed by them even if they were aware.

Interestingly, Mroczek et al. reported that the lowest overall quality of life and general health scores were found among residents of places where wind-farm developments were either at the stage of planning or under construction. (Mroczek et al. 2015). A possible explanation for this might be that those living near a WT, which were at the stage of planning or under construction, could have been irritated by the uncertain WT projects and had reported negative health problems. Residents may have been triggered by the negative attitude, subjective expectations or lifestyle. Kageyama’s findings showed that poor health symptoms were significant for noise sensitivity and visual annoyance but not significant for WTN indeed (Kageyama et al. 2016). Pawlaczyk-Łuszczyńska et al. revealed that general attitude towards WTs and attitude towards their visual impact was positively correlated (Pawlaczyk-łuszczyńska et al. 2014). In the study conducted by Jalali et al. (Jalali et al. 2016b), 45.9% of the participants had reported a negative attitude to WTs and 51.3% had positive or neutral attitude. Results of the Shepherd study presented that among the turbine group participants, 58.9% defined WTN as annoying. Also, the negative correlation between annoyance and self-rated health for both turbine and control groups was remarked (Shepherd et al. 2011). Klæboe et al. study results showed that noise annoyance depends strongly on separate non-acoustic factors. Noise sensitivity and visual and aesthetic factors had played a large role together with attitudes towards WPPs (Klæboe and Sundfør 2016).

General health symptoms and quality of life

The results of the studies (n = 11) examining the general health symptoms of people living around the WT farms are presented below.

In the first analyses of Pedersen’s research, distance to WT was negatively associated with unexplained fatigue, difficulty concentrating and dizziness. However, after the confounders (socioeconomic characteristics, noise-induced annoyance, anxiety about health, behavioural changes) were standardized in the modelling, the association between proximity to the WT and all physical symptoms was lost. The importance of this study lies in its being the very first cohort study in this field (Pedersen et al. 2009). In a study by Mroczek et al.(Mroczek et al. 2015), headache, anxiety, back pain, and severe fatigue were reported as daily complaints by those living near WTs. While the worst quality of life and general health outcomes were found among those living near WPPs that are still in the planning or installation phase, the highest quality of life was reported by those living near WPPs that are currently operating. Therefore, the presence of WTs around residential zones did not have a negative effect on the quality of life and health. It is also striking that WTN was never evaluated in this study. Michaud et al. (Michaud et al. Michaud, et al., 2016a) found no immediate association between general health and exposure to 46 dB WTN. However, it was indicated that sleep was significantly influenced by other factors, including, but not limited to, the use of sleep medication, other health conditions (including sleep disorders, higher total sleep time), and annoyance with blinking lights on wind turbines. In a study by Pawlaczyk-Łuszczyńska et al. (Pawlaczyk-łuszczyńska et al. 2014), a significantly higher number of participants who were disturbed by WTN reported dizziness, chest pain, sensation of ear pulsation, and irritability/stress. Negative attitude toward WT was correlated with the irritability. Shepherd et al.’s findings show that (Shepherd et al. 2011), the physical health and environmental health sub-scores of the quality-of-life scale were found to be lower in the group living within 2 km of the turbine site compared to those living within 8 km of the site. Another study of Nissenbaum (Nissenbaum et al. 2012) reported that participants in the high noise exposure group had significantly lower QOL compared to the low exposure group (Nissenbaum et al. 2012). The study conducted by Poulsen et al.(Poulsen et al. 2019) looked into incidence cases diagnosed with myocardial infarction (MI) and stroke, also taking into account all WT sites in operation in Denmark between 1980 and 2013. The 5-year exposure monitoring showed no increased risk in those exposed to outdoor night noise above 42 dB. The researchers stated that difference, if any, might not have been shown because disease incidence was not very high. MI and stroke due to WT have been hypothesized but no evidence exists to support this relationship. These results are likely to be related to nonspecific characteristics of these diseases or small exposure group numbers. Although these two diagnoses are classified under this title, it should be noted that they are non-specific. Smith et al. showed that the turbine group and control group did not differ in morning salivary cortisol levels (Smith Michael et al. 2020). Kageyama and Barry reported no evidence concerning the adverse effects of WTN on physical health on the basis of self-reported symptoms (Kageyama et al. 2016; Barry et al. 2018). The study of Crichton demonstrated that a sham information about sound produced by WT could create elevated concerns about the health effects of sound wind turbines (Crichton et al. 2014a;2014b).

Sleep disturbance

We found that ten of the articles under review investigated turbine-related sleep complaints and sleep quality issues.

According to the sleep quality-related results of the study by Pawlaczyk-Luszczyńska et al.(Pawlaczyk-łuszczyńska et al. 2014), 64% of the participants stated that they woke up rested while 26.3% had difficulty falling asleep. Difficulty falling asleep was significantly more common in the group reporting noise disturbance than in those who did not. Michaud et al. reported (Michaud et al. Michaud, et al., 2016a) that sleep disorders and sleep disturbances were not associated with the turbine induced noise. However, 36.8% of the group with 35 dB and above environmental noise exposure stated that they closed windows to avoid WTN before they went to sleep. The research team led by Michaud conducted another study (Michaud et al. 2016b) with the same population. In this study, sleep quality was not low in any of the groups classified according to WTN and no difference was found between the groups. It was found that WTN up to 46 dB did not affect sleep parameters. In the study conducted by Shepherd et al. (Shepherd et al. 2011), perceived sleep quality of the turbine group was reported to be lower than that of the control group. Jalali et al. (Jalali et al. 2016b) found that environmental noise-induced sleep disturbance started above 45 dB and 13.9% of the participants attributed their sleep disturbance to WTs. Sleep quality deteriorated significantly among those who could see the turbine from their house; those who had a negative attitude towards WT and those who had concerns about property values. It was found that the sleep-evaluating parameters (PSQI, ESS, ISI), which are the dependent variables of the study, deteriorated significantly after the installation of the wind farm. However, it was concluded that, contrary to expectations, the increase was not affected by the distance to the turbine, but by subjective considerations such as general attitude and visual annoyance. In the study of Kageyama et al.(Kageyama et al. 2016), the sex-age-adjusted ORs of insomnia were 5.55 for 40–45 dB group when compared with the reference (35 dB and below). There was a 4.17-fold increase in the risk for insomnia among those who expressed visual disturbance and a 24.4-fold increase in the risk for noise disturbance. As a result, it has been shown that sleep disorders, physical and mental disturbances can be affected by personal factors such as noise sensitivity and visual disturbance. In Jalali’s study (Jalali et al. 2016a), no significant change was found for PSG before and after the turbine site installation. Nonetheless, significant deterioration was noted in self-reported sleep quality and stress both at bedtime and in the mornings. Nissenbaum et al. reported (Nissenbaum et al. 2012) significant difference in sleep quality, daytime sleepiness, and mental health-related quality of life score with distance, between participants close to the turbine (375–1400 m) and far from it (3.3–6.6 km). In addition, the participants reported an improvement in their sleep quality when they moved away from the turbine site. Smith’s study(Smith Michael et al. 2020) found no objective outcome other than longer REM delay and less REM sleep in PSG evaluations during nights exposed to WTN. Nevertheless, participants reported lower sleep quality in the following morning after the WTN night compared to the morning after the control night. Additionally, they described increased fatigue, waking up more often, difficulty going back to sleep after waking up and poorer sleep than usual. It is noteworthy that self-reported sleep was reported to be worse than objective PSG results.. However it should be noted that this experiment was a laboratory sleep study which was conducted with simulated WTN. Also it is emphasized that there can be a stronger response to noise in laboratories than in the field which means that the findings of this study could potentially overestimate the effects of WTN on sleep. Barry et al. reported no statistically significant relationship between residential proximity to wind turbines and sleep quality which was examined with Pittsburgh Sleep Quality Index(Barry et al. 2018). In the findings of Mroczek study sleep problems was reported by 28.34% among the participants and the lowest QoL scores corresponded to recurrent problems with falling asleep as well as other health complaints(Mroczek et al. 2015).

General annoyance

Seven studies reported data on the annoyance related to WTs. In addition to general annoyance, studies also investigated annoyance towards several wind turbine features, including the following: blinking lights, vibrations, visual impacts and environmental pleasantness were included.

In the study of Jalali et al. (Jalali et al. 2016a) 45.9% of the participants stated a negative annoyance towards WT; similarly, in the study conducted by Pawlaczyk-Łuszczyńska et al. (Pawlaczyk-łuszczyńska et al. 2014), 20% of the participants expressed a negative annoyance towards WT both generally and visually. In this study, 63% of the general disturbance about the turbine could be explained by noise, general attitude to WTs, and sensitivity to landscape littering. In the study of Magari et al. (Magari et al. 2014), 90% of the participants were satisfied with their environment; 73% of them were not bothered by it despite being aware of the turbines when they were outside of their houses. Findings of Pedersen (Pedersen et al. 2009) showed that reported WT-related annoyance is correlated with the noise levels. The noticeability of the WTN is 78% within a range of 35–40 dB, is 87% within a range of 40–45 dB and is 92% above 45 dB. In the study of Barry et al. (Barry et al. 2018), annoyance outcome was associated with visual concerns, blinking lights, vibrations and also noise. It was reported that turbine-induced annoyance was associated with both distance and sound levels. It was revealed that annoyance could be reduced by about 20% per one kilometer of distance from a WT.

Michaud et al. (Michaud et al. Michaud, et al., 2016a) sleep (total sleep time) was significantly influenced by annoyance with blinking lights on WTs. In an experiment by Maffei et al., the visual annoyance reaction of a sample of the population was investigated with the general and WT pleasantness questions. The results showed that the factor distance (strongest) and the number (significant but small) and the colour influences the visual individual reactions (Maffei et al. 2013).

Shadow flicker effect

Annoyance related to shadow flickers and reflections from rotor blades was investigated by four articles.

In the study published by Pawlaczyk-ŁUszczyńska et al. (Pawlaczyk-łuszczyńska et al. 2014), 23.7% of the research group defined shadow flicker as annoying while 12.8% were annoyed about light reflection from the blades. Almost half of the participants stated that they noticed the shadow caused by the turbine. Those living at a distance of 400–800 m and 800–1200 m from the turbines were the most disturbed groups by the turbine-induced shadow flicker effect. Michaud et al. reported (Michaud et al. 2016b) that 7.8% of the participants reported disturbance caused by shadow flicker. On the other hand, in Magari’s study (Magari et al. 2014), people did not express any disturbance due to shadow or reflection caused by the WT while inside or outside the house. Barry’s findings showed that people who reported WT shadow flicker annoyance have no significant change in mean environmental quality of life score compared to those that did not report (Barry et al. 2018)

Discussion

This review demonstrates the 22 original articles about the possible effects of WTs on the health of people living near a wind farm. The main methodological characteristics containing the general quality and the bibliographic information of the studies were systematically evaluated. Among the reviewed articles, eleven were cross-sectional. Face-to-face interview data collection method was used in eight studies. Conflict of interest was stated in only two studies; seven studies did not mention limitations. Public support was reported in seven studies. The health outcomes, which were associated with WTs, were classified on the following topics: noise annoyance, risk perception, health symptoms and quality of life, sleep, annoyance, and flicker effect. Noise disturbance appears to be related to the dose of the noise exposure. We noted that the distance to the turbine had an effect on the attitudes towards WT, and this attitude and risk perception in their turn affected the frequency of the reported disturbances and symptoms. It seems reasonable to conclude that noise from WTs increases the risk of annoyance in exposed individuals in a dose–response relationship. Sources of environmental stress, planning uncertainties, and concerns about safety damage general health and quality of life as much as physical risk factors like noise do. Most studies show that any association between quality of life and health symptoms disappeared in multiple analyses where confounders such as personal perspective and psychological factors were adjusted. The reviewed articles indicated that sleep was negatively affected and sleep quality deteriorated due to WT exposure. It was noted that this change may be directly related to noise levels or it may be secondary to the disturbance. Studies show that individuals living in the vicinity of WT reported different levels of general annoyance. General annoyance about WTs is adversely affected by the noise level and sensitivity to noise. There were also complaints about shadow flicker and reflections around the wind farms.

There may be some possible limitations in this study. Only articles published in English language and between 2010 and 2020 were reviewed. Although the article evaluation process was conducted in the light of the PRISMA checklist, some points could have been missed.

Planning made in line with the opinions of the people living around the turbine and in accordance with the legal restrictions would have fewer risks for health. It should be noted that turbine noise reduces environmental satisfaction. Comprehensive analyses are needed at the installation phase in order to identify the most suitable locations with the highest potential for power generation and that will also respect people and nature.

Authors’ contributions

HH and AAT designed the study, AAT and HH prepared and analysed the data, and HH and AAT interpreted the data. AAT wrote the paper, with critical revisions from HH. All authors have read and approved the final version of this paper. All authors have accepted responsibility for the entire content of this manuscript and approved its submission

Ethical approval

The conducted research is not related to either human or animal use.

Informed consent

Informed consent is not applicable.

Data availability statement

The datasets used during the current study are available from the corresponding author on request.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported that there is no funding associated with the work featured in this article.