LAUSR

Background Numerous investigations have revealed significant relations between pain and fatigue in individuals with persistent pain conditions. However, the direction of influence between pain and fatigue remains unclear. Shortcomings of design and analytic approaches used in previous research limit the nature of conclusions that can be drawn about possible causal or directional relations between pain and fatigue. The present study investigated the temporal relation between changes in pain and changes in fatigue in individuals with musculoskeletal pain enrolled in a 10-week behavioral activation intervention. On the basis of previous findings, it was hypothesized that analyses would support a bi-directional relation between pain and fatigue. Methods The study sample consisted of 104 individuals with chronic musculoskeletal pain participating in a 10-week standardized rehabilitation intervention. Measures of pain intensity and fatigue were completed pre-, mid-, and post-treatment. The three-wave data panel permitted examination of the direction of influence between pain and fatigue through the course of the intervention. A random-intercept cross-lagged panel model (RI-CLPM) was used to examine the temporal relation between pain and fatigue. Results Consistent with previous research, cross-sectional analyses of pre-treatment data revealed significant correlations between measures of pain and fatigue. Significant reductions in pain and fatigue were observed through the course of treatment ( d  = 0.33 and d  = 0.66, p  < .001, respectively). RI-CLPM revealed that pain severity predicted later fatigue (pre to mid-treatment standardized path coefficient (β) = 0.55, p  = 0.02; mid to post-treatment β = 0.36, p  = 0.001); however, fatigue did not predict later pain severity. Conclusions Discussion addresses the processes that might underlie the temporal relation between pain and fatigue. Clinical implications of the findings are also discussed.