Skip to main content

Table 4 Association of different anti-Tat antibody isotypes with CD4+ T cell count and viral load

From: Association between different anti-Tat antibody isotypes and HIV disease progression: data from an African cohort

Associationa of duration of HIV infection and anti-Tat antibody isotype with CD4+ T cell counts (n = 88)
      Univariate    Multivariate  
Covariate Stratum n Mean CD4+ T cells count (cells/μL) IRR 95 % conf.int. p-value IRR 95 % conf.int. p-value
HIV infection (years) >3 69 403 1 1
≤3 19 501 1.24 (0.98 to 1.58) 0.079 1.10 (0.86 to 1.41) 0.443
IgG neg. 48 431 1 1
pos. 40 415 0.96 (0.76 to 1.22) 0.764 0.97 (0.77 to 1.21) 0.761
IgM neg. 43 353 1 1
pos. 45 491 1.39 (1.10 to 1.76) 0.007 1.34 (1.05 to 1.71) 0.019
IgA neg. 73 448 1 1
pos. 15 306 0.68 (0.51 to 0.91) 0.010 0.73 (0.55 to 0.95) 0.020
Associationa of duration of HIV infection and anti-Tat antibody isotype with Log10 VL (n = 83)
      Univariate    Multivariate  
Covariate Stratum n Mean viral load (Log10 copies/ml) IRR 95 % conf.int. p- value IRR 95 % conf.int. p-value
HIV infection (years) >3 65 4.68 1 1
≤3 18 4.26 0.91 (0.79 to 1.05) 0.203 0.96 (0.83 to 1.10) 0.553
IgG neg. 45 4.47 1 1
pos. 38 4.74 1.06 (0.96 to 1.16) 0.230 1.05 (0.96 to 1.15) 0.284
IgM neg. 40 4.84 1 1
pos. 43 4.36 0.90 (0.82 to 0.99) 0.027 0.92 (0.85 to 0.99) 0.037
IgA neg. 68 4.46 1 1
pos. 15 5.20 1.17 (1.08 to 1.26) <0.001 1.13 (1.06 to 1.21) <0.001
  1. IRR incidence rate ratio
  2. aPoisson regression with robust variance estimates was used to examine the association of Tat antibody isotypes with CD4+ T cell count and viral load. The table shows that the duration of HIV infection and anti-Tat IgG positivity had no significant influence on both parameters, whereas the presence of anti-Tat IgM and IgA were significantly associated with CD4+ T cell count and viral load. Anti-Tat IgM positivity was associated with higher CD4+ T cell counts and lower viral loads, whereas anti-Tat IgA positivity showed the opposite pattern. The fact that multivariate IRRs and p-values for both parameters are fairly similar to the univariate estimates suggest that these associations are independent of each other