Table 3

Association between loss of anti-human papillomavirus (HPV)-16 IgG and viral and demographic factors: univariate Cox proportional hazards regression analysis

No. lost IgG/ no. person-yearsRelative risk (95% confidence interval) for loss of IgGP
Age (yrs)
 <2024/65.11.43 (0.73–2.82)0.297
 White, Asian, and others25/84.51
 Hispanic7/13.31.80 (0.78–4.17)0.171
 African American5/23.30.71 (0.27–1.87)0.492
When anti-HPV-16 IgG was first detected
 During follow-up5/18.51
 At baseline32/102.61.39 (0.54–3.60)0.493
Initial absorbance value when anti-HPV-16 IgG was first detecteda
 ≤0.305533/38.416.54 (5.78–47.34)<0.0001
HPV types detected ≤ time tb
 HPV-16 related: 31, 33, 35, 52, or 585/16.72.07 (0.60–7.17)0.249
 Other types6/36.51.10 (0.34–3.63)0.870
 HPV negative21/32.74.05 (1.53–10.77)0.005
No. visits in which HPV DNA of any types was detected ≤ time tb
 27/22.11.69 (0.57–5.07)0.346
 13/21.20.91 (0.23–3.62)0.888
 021/32.73.88 (1.55–9.68)0.004
No. visits in which HPV-16 DNA was detected ≤ time tb
 0–132/99.91.35 (0.53–3.47)0.530
  • a Median initial absorbance among samples positive for IgG to HPV-16 virus-like particles.

  • b The independent variable was analyzed as a time-varying covariate. The sample sizes presented are number of subjects lost IgG/number of person-years at risk in the particular exposure category.