Skip to main content
. 2016 Jul 5;23(7):576–585. doi: 10.1128/CVI.00149-16

TABLE 2.

Association of HIV exposure and changes in infants' B-cell subset percentages and absolute counts during first 2 years of lifea

B-cell subset Effect parameter B-cell subset %
 B-cell subset absolute no.
Univariate linear regression
 Multivariate linear regression
 Univariate linear regression
 Multivariate linear regression
β coefficient (SE) P value β coefficient (SE) P value β coefficient (SE) P value β coefficient (SE) P value
B cells HIV exposure −0.139 (0.054) 0.012 −0.138 (0.055) 0.013 −0.063 (0.089) 0.482
Age −0.004 (0.004) 0.35 −0.004 (0.004) 0.378 −0.0126 (0.0067) 0.061
Naive B-cells HIV exposure −0.005 (0.017) 0.760 −0.019 (0.098) 0.850
Age −0.008 (0.001) <0.001 −0.020 (0.007) 0.004
Total MBCs HIV exposure −0.152 (0.110) 0.168 −0.054 (0.122) 0.657
Age 0.800 (0.062) <0.001 −1.430 (0.131) 0.000
Resting MBCs HIV exposure −0.211 (0.086) 0.015 −0.259 (0.065) <0.001 −0.191 (0.124) 0.124
Age 1.531 (0.093) <0.001 0.812 (0.045) <0.001 −1.351 (0.161) <0.001
Unswitched MBCs HIV exposure −0.337 (0.095) <0.001 −0.359 (0.081) <0.001 −0.287 (0.138) 0.039 −0.364 (0.128) 0.005
Age 0.620 (0.061) <0.001 0.629 (0.057) <0.001 −0.466 (0.067) <0.001 −0.493 (0.061) <0.001
Switched MBCs HIV exposure 0.131 (0.089) 0.140 0.201 (0.126) 0.111
Age 0.752 (0.046) <0.001 −1.34 (0.173) <0.001
Atypical MBCs HIV exposure 0.133 (0.081) 0.100 0.130 (0.116) 0.266
Age 0.009 (0.005) 0.103 −0.038 (0.019) 0.052
Activated B cells HIV exposure −0.022 (0.077) 0.780 −0.050 (0.106) 0.634
Age −0.954 (0.114) <0.001 −0.374 (0.064) <0.001
Plasmablasts HIV exposure 0.040 (0.144) 0.782 0.0370 (0.164) 0.823
Age 0.037 (0.009) <0.000 −0.335 (0.103) 0.001
a

Linear regression models were used to describe the estimated change in an infant's B-cell subset percentages and counts (β coefficients with standard errors). The impact of HIV exposure on B cells was determined at a significance level of a P value of <0.05. Subset analysis on naive B cells, total memory B cells, and plasmablasts was performed after adjusting for multiple testing using the Bonferroni correction at a significance level of a P value of <0.02, and further subgroup analysis within the memory B cells (resting, atypical, activated, switched, and unswitched memory B cells) was performed at a significance level of a P value of <0.005. Statistically significant values are shown in boldface type. The linear regression models were adjusted for clustering within an infant and, hence, accounted for inherent correlations between repeated measurements done on the same subject. MBC, memory B cell.