Wind farm noise woke sleepers less often than road traffic at the same volume

Objections to wind farms near housing routinely cite sleep. The evidence behind the claim has been mostly survey-based, which means it measures what people report about noise they know is there. A randomised laboratory trial published in the Journal of Sleep Research on October 22 takes the opposite approach: it plays the noise to sleeping people and measures their brain's response [s1].

The design

Sixty-eight adults spent a night in a sleep laboratory at Flinders University in Adelaide, South Australia, where they were exposed to repeated 20-second bursts of wind farm noise and road traffic noise [s1].

Exposures were delivered only after at least two minutes of established sleep, with at least 20 seconds between them [s1]. Pre-recorded samples of each noise type were reproduced at sound pressure levels of 30, 40 and 50 A-weighted decibels, in random order [s1].

The primary outcome was the probability of an electroencephalographically defined arousal event — a shift in EEG to faster frequencies lasting more than three seconds — following the onset of each exposure [s1]. Awakening responses, defined as frequency shifts lasting more than 15 seconds, were also assessed [s1].

Of the 68 participants, 62 had enough replicate exposures for analysis; their mean age was 49 (SD 20) and 35 were female [s1]. Effects of noise type, sound pressure level and sleep stage were modelled using mixed effects logistic regression [s1].

The result

Arousal probabilities were low overall, and particularly low in deep sleep [s1].

There was a significant interaction between noise type and sound pressure level (χ² = 13, p = 0.001) [s1]. At 40 dBA, arousal probability was 2.1% (95% CI 1.5-2.9) for wind farm noise against 3.2% (2.4-4.2) for road traffic noise (p = 0.016) [s1]. At 50 dBA it was 5.0% (4.0-6.2) against 8.6% (6.9-10.6) (p < 0.001) [s1].

Awakenings were infrequent — under 4% at 50 dBA — and showed similar effects [s1].

The authors' conclusion is deliberately narrow: acute wind farm noise onset is marginally less sleep disruptive than road traffic noise events of equivalent sound pressure level at 40 dBA and above [s1].

What "marginally" is doing in that sentence

The word is load-bearing. At 50 dBA the absolute difference between the two noise types is 3.6 percentage points in the probability that a given burst produces a brief EEG arousal [s1]. Both figures are small. Neither noise type reliably woke people up.

Two readings follow, and they are not in conflict. The first is that at the sound levels tested, neither noise source is a potent disruptor of established sleep in a laboratory. The second is that wind farm noise, which is the subject of the policy dispute, was the less disruptive of the two at matched levels.

What a laboratory night cannot tell you

This is an acute exposure study on a single night [s1]. It measures the response of a sleeping brain to a 20-second sound [s1] in a room where the participant knows they are being studied. It does not measure what happens across months of living near a wind farm.

Several things are missing by design. Sleep onset — whether noise makes it harder to fall asleep in the first place — is not the outcome here; the exposures were delivered only after sleep was established [s1]. Expectation and attitude towards the noise source, which the survey literature suggests matter substantially, are not modelled. And the participants were not people who live near wind farms, so habituation and sensitisation both fall outside the design.

The sound pressure levels tested — 30, 40 and 50 dBA [s1] — bracket what is typically measured in bedrooms near wind installations, but a laboratory reproduction is not the same acoustic environment as a house.

Why the objective measurement matters

The existing evidence base is largely subjective, and its quality is a known problem. A 2023 systematic review and meta-analysis of 15 articles found an overall prevalence of sleep disorders among residents near wind turbines of 34% (95% CI 0.22-0.47), with prevalence decreasing at greater distance (p = 0.010) and increasing with higher sound power level (p = 0.037) [s2].

But the same review concluded that the overall quality of research on the topic is poor, that the methods are often based on subjective assessments rather than validated questionnaires, and that no conclusions about causality can be drawn from the available data [s2]. It called for longitudinal designs and objective measurement [s2].

The October trial is an answer to the second of those requests, not the first.

What to watch

Whether field studies with in-home polysomnography or validated wearables replicate the laboratory finding; whether repeated-night exposure shows habituation or accumulation; and whether the sleep-onset question — distinct from arousal from established sleep — is addressed with the same objective methods.

This article describes laboratory research and is informational only. It is not medical advice, and it does not make a claim about the health effects of living near any particular installation.

Sources

• [s1] The Acute Effects of Wind Farm Versus Road Traffic Noise Onset on Electroencephalographically Defined Arousal From Sleep: Findings From an In-Laboratory Randomised Controlled Trial, Journal of Sleep Research, 35(3):e70227, published online 2025-10-22.
• [s2] Association between exposure to wind turbines and sleep disorders: A systematic review and meta-analysis, International Journal of Hygiene and Environmental Health, 2023;254:114273, published 2023-09-01.