Baseline Participant Characteristics are Associated With Engagement in a Movement-Based RCT for Chronic Low-Back Pain

Abstract

Purpose

Examine whether baseline participant characteristics predict engagement in a movement-based RCT for chronic low back pain (CLBP).

Design

Longitudinal study within an RCT.

Setting

Online.

Subjects

138 individuals with CLBP (18-79 years).

Intervention

Participants were randomized to a 12-week intervention of twice-weekly synchronous online yoga OR stretching/strengthening classes, and received 20-min pre-recorded home videos to complete daily.

Measures

Baseline questionnaires assessed sociodemographic, psychosocial, and pain factors (100% response rate). Engagement was operationalized as minutes of class attended + minutes of home videos completed.

Results

Bivariate correlations were computed between baseline variables and engagement. Three multivariate negative binomial generalized linear models (GLMs) were constructed to simultaneously examine predictors of engagement in the domains of sociodemographic, psychosocial, and pain-related factors. Greater engagement was significantly associated with greater baseline age, educational attainment, energy, and emotional well-being, and less emotion regulation difficulties, cannabis use problems, and pain interference (|rs| = .19-.33). In the domain specific GLMs, education (B = .491, P = .017) and cannabis use problems (B = −.048, P = .027) were the only significant predictors in the sociodemographic and psychosocial models, respectively. Neither pain interference nor pain severity were significant in the pain model.

Conclusion

Factors identified can inform strategies to increase engagement in movement-based CLBP interventions, potentially improving research validity and outcomes. Limitations include lack of racial diversity and not testing how engagement fluctuates throughout the intervention.