Special Poster Session 51st International Society for the Study of the Lumbar Spine Annual Meeting 2025

Introduction  

Exercise is often recommended and used as a treatment for chronic low back pain (cLBP).¹ While unsupervised exercise can improve pain and physical function, adherence to these regimens is often low.² It is unclear if exercise adherence trends exist across time and specific individuals. This study aims to identify patterns of unsupervised exercise frequency over one year and explore group differences in demographic, clinical, and psychosocial factors, and various treatment utilization.

Methods

We analyzed self-reported psychosocial, clinical, and behavioral data from a large observational cohort of individuals with cLBP, measured at 30-90 day intervals for 1 year (n=618). Group-based trajectory modeling classified participants based on exercise frequency and best model specifications. Baseline characteristics, patient-reported outcomes, and treatment utilization were compared using Kruskal-Wallis or chi-squared tests. Outcomes included PROMIS pain interference and physical function T-scores, Patient Health Questionnaire-2 (PHQ-2), Generalized Anxiety Disorder-2 (GAD-2), Tobacco, Alcohol, Prescription medication and other Substance use Tool (TAPS-1). Non-pharmacological treatment utilization was the number of different treatment types received (i.e., acupuncture, mind-body, diet, mental health, mindfulness, group exercise, physical/occupational therapy/chiropractic care).

Results

Our secondary analysis of a longitudinal study included 618 participants (out of 1000) averaging 59 years (SD=16.0), 37% males, and experienced cLBP for an average of 144.0 months (SD=144.5). Our analysis identified three groups of exercise frequency trajectories that remain stable throughout 1 year (High (5.0 days/week (SD=1.8)), Moderate (2.2 (SD=1.9), Low (0.6 (SD=1.3))(Figure 1).

The High-exercise group was mostly older, white, married, males, retired, with higher education, and a lower proportion of Medicaid recipients. This group reported the lowest proportion with substance use (TAPS 37.4%, χ²(2)=8.46, p=0.015) and the highest mean for non-pharmacologic treatment for back pain (1.43 (SD=1.22), χ²(2)=38.32, p<0.001).

The Low-exercisers were younger, non-White, unmarried, non-retired individuals with higher proportions of Medicaid and fewer college graduates. This group reported the lowest rates of non-pharmacological treatment use and the highest reported mean anxiety and depression scores (GAD-2 1.31 (SD=1.67), χ²(2)=8.55,  p=0.014, PHQ-2 1.30 (SD=1.63), χ²(2)=7.33, p=0.023).  The moderate group scored midway in between the high and low exercisers on most measures. Differences between groups reflected lifestyle factors though all groups had similar baseline pain and physical function scores (PROMIS pain interference, χ²(2)=1.72, p=0.426, PROMIS physical function, χ²(2)=4.28, p=0.118).

Discussion

Despite similar baseline pain and physical function scores, distinct differences in demographics, lifestyle, treatment use, and psychosocial outcomes were evident among the three exercise frequency groups. High-frequency exercisers who were older, wealthier, and predominantly white males, engaging in more frequent non-medical modalities also had lower rates of anxiety, depression, and substance use. Our findings allude to potential inequities in the uptake of beneficial non-pharmacologic therapies with the role of SES facilitating or inhibiting adherence to exercise treatment for cLBP.  Frequent exercise for cLBP may be disproportionately attainable for individuals with higher SES given the associated time, financial resources, and understanding of exercise benefits that were required. Future research should examine the causal relationships between exercise, psychological health, and treatment access and how addressing these potential barriers may improve exercise adherence.