Table 2

Multiple linear regression models estimating the average change in CD4 cell count per month (CD4 slope). Separate models analyse the first (0 – 16 weeks) and second (16 – 48 weeks) phases of immune recovery during ART.

Phase 1 (0–16 weeks) CD4 slope (n = 596)

Phase 2 (16–48 weeks) CD4 slope (n = 311)1

Mean

(95% CI)

P value

Mean

(95% CI)

P value


Age (years)

<30

1.0

1.0

30–39

-2,27

(-6.71,2.17)

0.32

-2.18

(-5.98, 1.63)

0.26

>40

-2.17

(-7.87, 3.54)

0.46

-7.31

(-12.11, -2. 50)

0.003

Sex

Female

1.0

1.0

Male

-3.60

(-8.22, 1.02)

0.13

-0.94

(-4.98, 3.10)

0.65

Baseline CD4 coimt (cells/μl)

>150

1.0

1.0

100–149

3.85

(-1.63, 9.33)

0,17

-2.17

(-7.09, 2.76)

0.39

50–99

3.53

(-1.84, 8.91)

0.20

0.93

(-4.22, 6.08)

0.72

<50

2.34

(-3.31, 7.98)

0.42

8.55

(2.39, 14.71)

0.007

WHO clinical stage

1 & 2

1.0

1.0

3

1.79

(-3.33, 6.91)

0.49

-0.70

(-5.40, 4.00)

0.77

4

0.55

(-5.39, 6.49)

0.86

-1.76

(-6.93, 3.41)

0.50

Baseline viral load (copies/ml)

<5 log10

1.0

1.0

>5 log10

11.14

(7.10, 15.18)

<0.001

2.21

(-1.37, 5.80)

0.23


1 The Phase 2 model included as covariates all variables shown, as well the phase 1 CD4 cell slope, the CD4 cell count at 16 weeks and the viral load at 16 weeks (see text).

Lawn et al. BMC Infectious Diseases 2006 6:59   doi:10.1186/1471-2334-6-59

Open Data