Final analytic model of pain interference using two-part regression modeling

Variables within hierarchical modelsFirst part—logistic regression
(n = 701)
Any pain interference (binary)
Second part—OLS regression
(n = 459)
Magnitude of pain interference
Overall model stats
ORSE95% CIPPseudo R2bSE95% CIPR2AICLR test
Base covariates< 0.00010.222< 0.00010.1833,751Base
Age1.020.011.00–1.040.02–0.100.03–0.16––0.05< 0.001
Gender (ref. female)
    Male0.650.180.38–1.110.121.090.93–0.74–2.920.24
    Transgender, non-binary, gender non-conforming, or other1.210.880.29–5.060.80–2.121.91–5.86–1.630.27
Education (ref. ≤ HS)
    Associates degree or some college0.130.110.03–0.650.01–0.861.23–3.27–1.550.47
    Bachelor’s degree0.090.070.02–0.44< 0.01–2.881.22–5.26––0.500.02
    Master’s degree or above0.080.070.02–0.40< 0.01–3.431.23–5.83––1.03< 0.01
MS subtype (ref. RRMS)
    PPMS1.440.690.57–3.670.440.331.37–2.37–3.020.81
    SPMS1.570.540.81–3.070.182.580.910.80–4.36< 0.01
    PRMS1.351.290.21–8.80.755.712.361.08–10.35< 0.02
    Unsure1.130.630.38–3.350.822.811.70–0.52–6.140.10
DMT (ref. no therapy)
    First line1.490.430.85–2.610.17–0.630.91–2.39–1.140.49
    Second line2.220.651.25–3.95< 0.01–1.320.86–3.01–0.380.13
Pain med count2.310.231.90–2.80< 0.0011.020.240.56–1.48< 0.001
Childhood stressors< 0.00010.245< 0.00010.2153,719< 0.0001
Child stressor count1.190.220.83–1.710.34–0.580.54–1.63–0.470.28
Child stressor severity0.990.040.91–1.070.730.140.12–0.10–0.380.24
Adult stressors< 0.00010.258< 0.00010.2673,685< 0.0001
Adult stressor count1.040.021.00–1.080.070.060.05–0.04–0.160.24
Adult stressor severity1.000.010.98–1.020.920.040.02–0.01–0.080.10

≤ HS: high school equivalency or below; PPMS: primary progressive MS; PRMS: progressive-relapsing MS; ref.: reference; SPMS: secondary progressive MS. Blank cells indicate not applicable to individual variables