Clarifying the Nature of Pain-Related Anxiety: Implications for Assessment and Treatment of Chronic Musculoskeletal Pain

Abstract

Pain-related anxiety and anxiety sensitivity (AS) are important constructs in fearanxiety- avoidance models of chronic pain (Asmundson, P. J. Norton, & Vlaeyen, 2004). Pain-related anxiety (McCracken & Gross, 1998) includes dimensions of cognitive anxiety (e.g., concentration difficulties as result of pain), behavioural avoidance, fearful thinking about pain, and physiological reactivity to pain (e.g., autonomic arousal, nausea). AS (Reiss, Peterson, Gursky, & McNally, 1986) is the trait tendency to fear the physiological sensations of anxiety due to the belief such sensations signal imminent harm. Evidence suggests an association between AS and pain-related anxiety (e.g., Muris, Schmidt, Merckelbach, & Schouten, 2001; P. J. Norton & Asmundson, 2003); however, the nature of this relationship remains unclear. An overlapping but empirically distinct relationship has been suggested (Carleton, Abrams, Asmundson, Antony, & McCabe, 2009) but there is also evidence pain-related anxiety may be a manifestation of AS (Greenberg & Burns, 2003). The current study sought to assess the posited view that pain-related anxiety may be an expression of AS. An experimental design was used in an attempt to extend the findings of Greenberg and Burns (2003) with a non-clinical analogue sample. Participants were healthy adults (N = 61, 62% women, M age = 31, SD = 11.45) who completed measures of pain-related anxiety, AS, social anxiety, fear of negative evaluation, and general negative affectivity (i.e., depression, trait anxiety). They underwent a pain induction task intended to elicit pain-related anxiety and a mental arithmetic task intended to elicit social-evaluative anxiety. Data gathered at baseline, during, and post-experimental tasks included (a) cardiovascular variables to provide indices of anxious arousal; (b) self-report measures of pain-related anxiety, social evaluative anxiety, and general negative affectivity; and (c) behavioural performance measures (i.e., correct answers on the mental arithmetic task, pain tolerance). Two hypotheses were tested: 1. Consistent with the view that pain-related anxiety may be a manifestation of AS, it was hypothesized that a measure of pain-related anxiety (i.e., Pain Anxiety Symptoms Scale-20[PASS-20]; McCracken & Dhingra, 2002) would significantly and substantively predict scores on post-task dependent measures for both the pain-related anxiety and social-evaluative anxiety induction tasks in regression models while controlling for effects of general negative affectivity; 2. It was hypothesized that the predictive effects of pain-related anxiety (PASS-20) on dependent measure scores would be accounted for by scores on a measure of AS (Anxiety Sensitivity Index-3 [ASI-3]; Taylor et al., 2007) in regression models. Neither of these hypotheses was supported. For the first hypothesis, results revealed that PASS-20 scores predicted positive variance in only the pain induction post-task measure of current painanxiety. Contrary to prediction, the PASS-20 did not account for variance in any of the mental arithmetic task dependent measures. For the second hypothesis, the results similarly failed to reject the null hypothesis. Despite exhibiting a high degree of correlation with the PASS-20, ASI-3 scores failed to account for positive variance in either the pain induction or mental arithmetic post-task dependent measures. Results indicated that AS was not associated with pain-related anxiety in a sample of participants not reporting current pain. These findings may lend support to the view that the apparently robust relationship observed between AS and pain-related anxiety among persons with chronic pain, may, in part, be a consequence of a persistent pain experience.