The role of self-efficacy and control beliefs in response to a multimodal headache intervention: results from a prospective observational study with a waiting-list comparator

Primary headache disorders, particularly migraine, rank among the leading causes of disability worldwide and impose substantial social and economic burden. Multimodal headache treatment (MMHT) combines pharmacologic, physiotherapeutic and cognitive-behavioural interventions and has shown benefit in high-burden patients, yet the psychological factors associated with sustained improvement remain unclear. Self-efficacy (SE) supports active coping, whereas the chance-related multidimensional health locus of control (CMHLC-C) reflects belief in uncontrollable outcomes. We hypothesised that these control beliefs are modifiable through intervention and that their baseline levels are associated with subsequent improvement in headache burden. Adults with primary headache disorders were enrolled in a prospective observational study and participated in a one-week MMHT at a tertiary neurological day clinic. Headache impact (HIT-6; primary endpoint), monthly headache days (MHD), and highest headache pain severity (HHPS) were measured three months before treatment (waiting-list baseline), at the start and end of MMHT, and at 3-, 6-, and 9-month follow-up. The waiting-list period served as a within-person pre-treatment comparator for spontaneous change. Baseline headache-management SE and MHLC-C subscales were tested as predictors of longitudinal HIT-6 trajectories using a generalized linear mixed model with visit as a repeated factor. Sixty-five patients were included in the analytic cohort (84.6% female; mean age 40.6 ± 13.3 years; 89.8% migraine). HIT-6 decreased from 63.2 at the waiting-list baseline to 59.0 at nine months (p<.001). MHD was reduced from 17.9 ± 8.0 to 12.2 ± 7.5 days, and 44% experienced a ≥ 30% reduction. HHPS also decreased throughout follow-up. Mean SE increased significantly during MMHT and early follow-up, whereas CMHLC-C showed only small changes over time. The mixed model was significant (F(2,165) = 7.25, p<.001). Higher baseline SE predicted larger reductions in HIT-6 (β=−0.55, t(165) = − 2.62, p=.010), whereas greater belief in chance predicted smaller reductions (β = 0.85, t(165) = 2.80, p=.006). In this prospective observational study with a waiting-list comparator, MMHT was associated with sustained reductions in headache impact, headache frequency, and pain severity. Baseline self-efficacy and chance-related control beliefs independently predicted treatment response and may represent clinically relevant stratification targets in personalised headache care.