Wind turbines and idiopathic symptoms: The confounding effect of concurrent environmental exposures

• Concerns have been raised about the potential health effects of wind turbines.
• No study on wind turbines and health has controlled for other co-exposures.
• Other environmental co-exposures are confounders.
• After controlling for them, there was no wind turbines-symptoms association.
• Associations reported in the literature may be due to confounding bias.

Whether or not wind turbines pose a risk to human health is a matter of heated debate. Personal reactions to other environmental exposures occurring in the same settings as wind turbines may be responsible of the reported symptoms. However, these have not been accounted for in previous studies. We investigated whether there is an association between residential proximity to wind turbines and idiopathic symptoms, after controlling for personal reactions to other environmental co-exposures. We assessed wind turbine exposures in 454 residences as the distance to the closest wind turbine (Dw) and number of wind turbines < 1000 m (Nw1000). Information on symptoms, demographics and personal reactions to exposures was obtained by a blind questionnaire. We identified confounders using confounders' selection criteria and used adjusted logistic regression models to estimate associations. When controlling only for socio-demographic characteristics, log10Dw was associated with “unnatural fatigue” (ORadj = 0.38, 95%CI = 0.15–1.00) and “difficulty concentrating” (ORadj = 0.26, 95%CI = 0.08–0.83) and Nw1000 was associated with “unnatural fatigue” (ORadj = 1.35, 95%CI = 1.07–1.70) and “headache” (ORadj = 1.26, 95%CI = 1.00–1.58). After controlling for personal reactions to noise from sources different from wind turbines and agricultural odor exposure, we did not observe a significant relationship between residential proximity to wind turbines and symptoms and the parameter estimates were attenuated toward zero. Wind turbines-health associations can be confounded by personal reactions to other environmental co-exposures. Isolated associations reported in the literature may be due to confounding bias.