Table 3.

Factors affecting IgG-S titers by multivariate linear regression for the entire cohort with or without adjustment

	Without adjustment				With adjustment
Variable	Estimate	SE	p		Estimate	SE	p
(Intercept)	10.85	0.272	< 0.001		10.32	0.711	< 0.001
Time lag, month	-0.356	0.036	< 0.001		-0.304	0.058	< 0.001
HTLV-1 infection	-0.465	0.188	0.014		-0.501	0.179	0.006
Age	-0.020	0.006	0.001		-0.009	0.012	0.450
Male gender	-0.081	0.158	0.610		-0.213	0.200	0.288
Higher BMI	-0.131	0.257	0.610		-0.053	0.252	0.833
Drinking habit	-0.066	0.159	0.679		0.136	0.142	0.338
Smoking habit	-0.182	0.211	0.392		-0.157	0.161	0.332
Presence of diabetes, hypertension, or dyslipidemia	0.061	0.147	0.678		-0.051	0.146	0.728
COVID-19 vaccination history involving different types of combinations	-0.161	0.155	0.300		-0.311	0.161	0.056

The propensity score method using overlap weights was employed to adjust for confounding

Abbreviations: HTLV-1, human T-lymphotropic virus type 1; BMI, Body Mass Index; SE, standard error; p, p value. Higher BMI is defined as “BMI > = 30.” Drinking habit is defined as “drinking alcohol more than 3 days per week.” Smoking habit is defined as “still have a habit”