Resistance to extinction in an odor–20% CO 2 inhalation paradigm: Further evidence for a symptom learning account of multiple chemical sensitivity Ann Meulders, Stien Fannes, Ilse Van Diest, Steven De Peuter, Debora Vansteenwegen 1 , Omer Van den Bergh ⁎ Department of Psychology, University of Leuven, Leuven, Belgium Received 18 January 2008; received in revised form 25 February 2009; accepted 9 March 2009 Abstract Objective: Multiple chemical sensitivity (MCS) is characterized by a variety of symptoms in response to nontoxic concentrations of chemicals. To further test a laboratory model of MCS based on symptom learning, we used a stronger respiratory challenge in this study than in previous studies to induce symptoms (20% CO 2 - enriched air, unconditioned stimulus). Additionally, extinction of learned symptoms was investigated. Methods: Diluted ammonia (AM) and butyric acid (BA) served as conditioned (odor) stimuli (CSs). In a learning phase, healthy participants received three breathing trials of BA mixed with CO 2 and three trials of AM mixed with regular air, or the reversed combination. In a test phase, the same trials were administered without CO 2 . Immedi- ately after odor onset, participants rated their anxiety and expectancy to experience symptoms during that trial. Self-reported symptoms were assessed after each trial; respiratory behavior was measured throughout the experiment. Results: Only participants who learned to correctly anticipate symptoms (learning phase) reported elevated symptoms in response to the CS+ odor (test phase). Anticipation was more likely correct and learned symptoms were more elevated when BA had been previously associated with CO 2 . Across test trials, anticipatory anxiety diminished, but learned symptoms did not. Participants failing to learn the prearranged contingencies reported overall more symptoms and anxiety. Conclusions: Strong respiratory challenges impede extinction of learned symptoms. Conscious expectancy, which may be modulated by odor quality, determines whether learned symptoms develop in response to a specific odor or to the general context. The results further support a symptom learning account of MCS. © 2010 Elsevier Inc. All rights reserved. Keywords: Multiple chemical sensitivity; Associative learning; Extinction; Online US expectancy; Odors; CO 2 Introduction Multiple chemical sensitivity (MCS) is an ill-defined clinical syndrome characterized by a wide variety of symptoms in different organ systems in response to chemical substances far below toxic concentrations [1]. Typical symptoms include fatigue, dizziness, poor memory, muscle and joint pain, shortness of breath, and anxiety [2,3] and largely overlap with those of other controversial functional syndromes such as fibromyalgia and chronic fatigue syndrome. MCS patients, however, distinctively attribute their symptoms to chemicals in their environment that typically do not elicit symptoms in the majority of people. Numerous explanations, both physiological and psycholo- gical, have been advanced, but none of these accounts have received unequivocal support (see Refs. [2,4,5] for reviews). We developed a laboratory model of MCS, based on associative learning, and collected a substantial body of evidence showing that self-reported symptoms can be learned in response to chemicals (i.e., odors) presented in harmless concentrations [6–8]. The typical learning paradigm administers Journal of Psychosomatic Research 68 (2010) 47 – 56 ⁎ Corresponding author. Department of Psychology, Tiensestraat 102, B-3000 Leuven, Belgium. Tel.: +32 0 16 326058; fax: +32 0 16 325923. E-mail address: omer.vandenbergh@psy.kuleuven.be (O. Van den Bergh). 1 Debora Vansteenwegen is a postdoctoral researcher at the Research Foundation—Flanders, Belgium (FWO). 0022-3999/09/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jpsychores.2009.03.009