Health Canada, the country’s health department, is preparing to launch a study of health effects among residents near wind farms.  The study as currently planned will be based on interviews and physiological measurements of 2000 people living near wind farms of 8-12 turbines.  Each participant will fill out a self-reported health survey, and will be tested for stress hormones, blood pressure, and sleep patterns.  Study subjects will live from under 500 meters from turbines to over 5km.

The research design and methodology has just been made public, and is open for public comment until August 8.  Among the key methods outlined, with my initial thoughts in brackets:

• Sleep patterns will be monitored using wrist-worn devices for seven consecutive nights. [Will this be long enough to capture sleep-disruption trends, if they exist?  It would be good to ask study participants who self-report sleep disruption whether the week chosen for testing was representative of their worst weeks, average weeks, or below-average weeks, in terms of sleep quality.  A pilot study will examine the usefulness of adding a sleep diary to the full study protocol; this should be encouraged as a way to assess sleep patterns over a longer period of time, ideally including seasonal differences.]
• Sound levels will largely be modeled based on measured sound levels near the turbines, including sound into the infrasonic range; these models will be validated in the field at distances of up to 5km. [It would be good to know the full range of frequencies that will be modeled and measured.  In addition, medium- to long-term validation measurements in the field would be useful, in order to capture a better sense how often worst-case noise periods may occur; such events may be relatively rare or seasonal but be important elements in community response, especially stress responses.  Models to be utilized should be based on recent studies that have found lower frequency elements of wind turbine noise often attenuate at a lower rate than higher frequencies; this is especially important in considering any possible effects of audible low frequency sound at distances of a kilometer/half mile or more.)
• Sampling out to 5km (3.1 miles) will allow researchers to generate a dose-response curve based on the sound levels of the turbines.  [5km is probably a decent distance to use for an effective control, in that turbines are nearly always inaudible at such a distance.  I would hope that the study design can assure that there are enough subjects at close range, especially within 1km (.6 miles) and 2km (1.25 mi). to be sure that any reported and measured health effects will be statistically significant.  While I appreciate the need to have statistical significance at all distances, “wasting” too many subjects at greater distances could make it more difficult to be sure of any effects found at the distances where they appear to be more common, and at which there is apt to be a greater range and complexity of responses.]
• [While a dose-response relationship is a foundation of many health effects, there are many other factors that tend to make individuals more or less susceptible to whatever health stressor is being studied.  It would be helpful to include assessment of some of the factors that could be contributors to a health effect from wind farm noise, including noise sensitivity, pre-existing vestibular issues, and susceptibility to motion sickness, among others.]