Beyond headache frequency: How acute treatment response shapes migraine-related disability-Insights from a joint stratification framework: A cross-sectional study

The aim of this study was to examine whether joint profiling of headache frequency and acute treatment response identifies clinically meaningful patient subgroups with different disability burden and treatment needs, and whether per-attack acute treatment response provides disability-relevant information beyond monthly headache days and other measured covariates. Headache frequency guides migraine treatment escalation, yet the quality of acute treatment response may represent an equally important but underassessed determinant of disability.

This cross-sectional study enrolled 404 consecutive patients with migraine at a tertiary headache center. Acute treatment response was measured with the 12-item Chinese Migraine Treatment Optimization Questionnaire (mTOQ-C12) (score 0-12; ≤9 indicating suboptimal response). Disability and quality of life were assessed with the Migraine Disability Assessment Scale and the Migraine-Specific Quality of Life Questionnaire version 2.1. Patients were stratified into four subgroups by cross-tabulating binary monthly headache days (≥8 vs. <8 days per month) and binary treatment response. Associations were evaluated after adjustment for demographic, headache-related, and psychological covariates.

The four subgroups showed significant overall differences in disability scores (median range = 4.0-20.3 points; p < 0.001) and quality-of-life scores (median range = 68.7-80.7; p < 0.001). Among low-frequency patients, suboptimal responders showed threefold higher disability than adequate responders despite identical median attack frequency (both 3.0 days per month; median disability score 12.0 vs. 4.0; Bonferroni-adjusted p < 0.05). Notably, low-frequency/suboptimal response patients showed covariate-adjusted disability similar to high-frequency/adequate response patients (adjusted mean difference = -0.97 points; 90% CI = -4.70 to +2.76; equivalence p = 0.038 within ±5-point bound), despite a fivefold frequency difference (3.0 vs. 15.0 days per month). In an exploratory low-frequency subgroup analysis, the monthly headache days (MHD)-disability association was numerically stronger among suboptimal responders (B = +0.21, 95% CI = +0.09 to +0.33, p < 0.001) and was substantially attenuated among adequate responders (B = +0.12, 95% CI = -0.03 to +0.27, p = 0.107), but the between-stratum difference was not significant (ΔB = +0.10, 95% CI = -0.07 to +0.28, p = 0.255). After full covariate adjustment, the mTOQ-C12 remained associated with ln (Migraine Disability Assessment Scale+1) (B = -0.08, 95% CI = -0.12 to -0.04, p < 0.001; ΔR² = 0.03; standardized β =-0.17). In the sequential block-entry regression, the standardized contribution of mTOQ-C12 was lower than the Patient Health Questionnaire-9 (PHQ-9) and broadly comparable to MHD. In exploratory machine-learning analyses, mTOQ-C12 ranked second by mean Shapley additive explanations after PHQ-9, whereas its top three selection probability ranked third after PHQ-9 and MHD.

Poorer acute treatment response was associated with greater migraine-related disability beyond headache frequency and other measured covariates. Joint profiling of headache frequency and treatment response identified clinically meaningful disability patterns not captured by headache frequency alone, supporting routine assessment of acute treatment response as a complementary clinical dimension in migraine care.