Table 4 Statistically significant univariate regression models for the strength of migraine assessments and multivariate models with adjustment by covariates: age and gender.

Variable	Univariate regression models					Multivariate regression models
Variable	BETA	95% LCI BETA	95% UCI BETA	p	R2	BETA	95% LCI BETA	95% UCI BETA	p	R2
Logistic regression models for MIDAS-degree
CD3 + CD57 + (% of CD3 +)	0.92	0.84	1.00	0.058	0.13	0.94	0.85	1.02	0.144	0.17
CD19 non-switched memory abs	1.02	1.00	1.04	0.095	0.09	1.01	0.99	1.04	0.184	0.16
OLS regression models for MIDAS (values from 0 to 100)
CD4 + T_EM (% of CD4 +)	0.97	0.93	1.00	0.068	0.08	0.98	0.94	1.02	0.243	0.19
Logistic regression models for HIT-6
CD4 + T_EM (% of CD4 +)	0.82	0.69	0.93	0.01	0.33	0.84	0.70	0.95	0.02	0.38
CD3 + CD57 + (% of CD3 +)	0.92	0.85	1	0.062	0.13	0.9	0.8	1	0.058	0.28
Treg (CD4) (% of CD4 +)	0.76	0.55	1	0.065	0.13	0.74	0.5	1.01	0.086	0.26
CD4 naïve (% of CD4 +)	1.07	0.99	1.16	0.098	0.11	1.04	0.96	1.14	0.398	0.18

BETA is the estimated regression coefficient. For logistic regression models, the odds ratio is reported (exponentiated regression coefficient). 95% LCI/UCI BETA is 95% lower and upper confident interval estimates. R2 is the coefficient of determination for OLS model and pseudo coefficient of determination for logistic models.

Quick links

Search