Table 2. Summary of the regression analyses for the three cold sore serostatus phenotypes (N=622 subjects).

Factor	Intercept	Sex	Ethnicity	Age	Hap 2	Hap 3	Hap 4	Hap 5/6
Response variable=serostatus (EI >1.1)
Effect size	−1.102	0.226	0.093	0.044	0.142	0.003	−0.062	0.492
Conventional s.e.	0.222	0.128	0.183	0.006	0.164	0.205	0.181	0.264
Conventional P value	<0.001	0.078	0.611	<0.001	0.387	0.99	0.73	0.062
Bootstrap s.e.	0.304	0.183	0.271	0.008	0.164	0.213	0.182	0.256
Bootstrap P value	<0.001	0.207	0.723	<0.001	0.385	0.987	0.734	0.044
Response variable=serostatus (EI >2.5)
Effect size	−1.253	0.225	0.216	0.044	0.065	−0.096	−0.16	0.317
Conventional s.e.	0.217	0.126	0.181	0.005	0.161	0.202	0.179	0.253
Conventional P value	<0.001	0.074	0.232	<0.001	0.689	0.634	0.371	0.21
Bootstrap s.e.	0.296	0.179	0.27	0.008	0.162	0.21	0.177	0.248
Bootstrap P value	<0.001	0.2	0.403	<0.001	0.69	0.644	0.363	0.191
Response variable=EI
Effect size	1.933	0.387	0.11	0.083	0.196	0.122	−0.357	0.672
Conventional s.e.	0.432	0.256	0.37	0.01	0.33	0.415	0.368	0.5
Conventional P value	<0.001	0.131	0.766	<0.001	0.552	0.769	0.331	0.179
Bootstrap s.e.	0.567	0.359	0.512	0.013	0.335	0.44	0.363	0.475
Bootstrap P value	<0.001	0.281	0.834	<0.001	0.559	0.784	0.323	0.161

Subjects were defined as HSV1 seropositive if the EI was >1.1 (manufacturer’s recommendation) or EI >2.5 (more stringent definition). The regression analysis was also performed on the continuous variable of EI. Significant values are in bold.

Abbreviations: EI, enzyme-linked immunosorbent assay index; Hap, haplotype.