Importance: Low back pain (LBP) is influenced by interrelated physical, psychological, and social factors. However, most treatments focus on symptom reduction without addressing the underlying biopsychosocial needs of patients.
Objective: To determine the effectiveness of spinal manipulation and clinician-supported biopsychosocial self-management vs medical care for adults with increased risk of chronic disabling LBP.
Design, setting, and participants: This 2 × 2 factorial randomized clinical trial enrolled participants in 3 research clinics at the Universities of Minnesota and Pittsburgh from November 2018 to May 2023; final follow-up was in June 2024. Adults with acute or subacute LBP at moderate to high risk of chronicity based on the STarT Back tool were randomized to 1 of 4 groups, with interventions lasting up to 8 weeks. Statistical analysis was conducted from November 2024 to June 2025.
Interventions: Spinal manipulation therapy (n = 201), supported self-management (n = 305), or combined supported self-management with spinal manipulation (n = 193) compared with guideline-based medical care (n = 301). Physical therapists and chiropractors provided spinal manipulation and supported self-management.
Main outcomes and measures: The 2 primary outcomes averaged over a follow-up of 1 year were monthly low back disability (Roland-Morris Disability Questionnaire) and weekly pain intensity (numerical rating scale). Secondary analysis examined the proportion of participants achieving a 50% or higher reduction in the primary outcome measures.
Results: Among the 1000 participants randomized (mean [SD] age, 47 [16] years; 58% female), 93% completed the trial. The omnibus test for differences across the 4 treatment groups was statistically significant for disability (P = .001; supported self-management, 4.7; spinal manipulation, 5.5; combined supported self-management with spinal manipulation, 4.8; medical care, 5.9) but not pain intensity (P = .16; supported self-management, 2.8; spinal manipulation, 3.0; combined supported self-management with spinal manipulation, 2.8; medical care, 3.0). Averaged over 12 months, LBP disability was significantly lower compared with medical care for supported self-management (mean difference, -1.2 [95% CI, -1.9 to -0.5]) and supported self-management with spinal manipulation (mean difference, -1.1 [95% CI, -1.9 to -0.3]) but not spinal manipulation alone (mean difference, -0.4 [95% CI, -1.2 to 0.4]). Group differences in pain intensity were not statistically significant; point estimates ranged from -0.2 to 0. Both supported self-management groups had higher proportions of patients achieving a 50% or greater reduction in disability (supported self-management, 67%; spinal manipulation, 54%; combined supported self-management with spinal manipulation, 65%; medical care, 54%).
Conclusions and relevance: For patients with acute or subacute LBP at increased risk of chronic disabling LBP, clinician-supported biopsychosocial self-management showed statistically significant but small reductions in disability, but not pain, vs medical care over 1-year follow-up, and spinal manipulation alone showed no significant difference for either outcome.
Trial registration: ClinicalTrials.gov Identifier: NCT03581123.