Mediators of the effect of chiropractic care on 12- and 52-week outcomes for U.S. active-duty military personnel with low back pain: secondary analysis of a clinical trial

The aim of this secondary analysis is to examine pain and non-pain factors as mediators of the effect of chiropractic care on 12- and 52-week outcomes from a clinical trial subsample of U.S. active-duty military personnel. Beginning in November 2014, the final 154 participants enrolled in a clinical trial, were asked to provide outcome data for 52 weeks. Data collection concluded in February 2016. We used natural effect models to evaluate 6-week change in the Roland-Morris Disability Questionnaire (RMDQ) and Numerical Pain Rating Scale (NPRS) as mediators of the effect of chiropractic care on 12- and 52-week outcomes of fatigue, sleep disturbance, and social role. We also evaluated non-pain factors as mediators of the effect of chiropractic care on pain outcomes. Models were adjusted by trial allocation minimization variables and by baseline values of the mediators and outcomes and results are reported as between-group differences on the outcome scales with 95% confidence intervals. One hundred forty-four participants were included in the analysis. Participants had an approximate mean age of 33 years, were 78% male, and 63% white race. Except for NPRS, which demonstrated negligible difference at 52 weeks, total effects favored the group receiving chiropractic care. At 52-weeks the indirect effects through RMDQ were fatigue: 0.64 (95%CI −0.14 to 1.60), sleep: 0.82 (0.15 to 1.64), and social: 0.75 (0.12 to 1.65), mediating 44%, 17%, and 33% of the total effects. Indirect effects through non-pain factors on 52-week RMDQ were fatigue: 0.23 (−0.39 to 0.74), sleep: 0.20 (−0.32 to 0.80), and social: 0.38 (−0.06 to 0.94), mediating 12%, 10%, and 19% of the total effects. Improvement in pain intensity and pain-related disability during chiropractic care were drivers of 12- and 52-week sleep disturbance and social role outcomes for U.S. active-duty military. Mediation by fatigue, sleep disturbance, and social role of longer-term pain outcomes was small to negligible. A larger sample size and study of differing patient populations is needed to better understand how to tailor chiropractic care to target specific intermediates and achieve better patient outcomes. The trial was prospectively registered on clinicaltrials.gov (NCT01692275).