Prevalence and correlates of cytopenias in HIV-infected adults initiating highly active antiretroviral therapy in Uganda

Table 3 Factors associated with the presence of any cytopenia

Results of uni- and multivariable log-binomial regression models
Covariate	Univariable (n = 400)		Multivariable‡ (n = 397*)
	PR (95% CI)	p value	PR (95% CI)	p value
Sex
Male	1		1
Female	1.21 (1.01 to 1.43)	0.035	1.33 (1.12 to 1.59)	0.002
CD4 count cells/μl
<50	1		1
50 to < 200	0.73 (0.63 to 0.85)	<˂0.001	0.75 (0.64 to 0.88)	<0.001
200 to < 350	0.70 (0.57 to 0.86)	0.001	0.74 (0.59 to 0.92)	0.007
≥ 350	0.79 (0.52 to 1.19)	0.256	0.77 (0.50 to 1.19)	0.245
BMI
under weight	1		1
normal	0.79 (0.67 to 0.95)	0.011	0.82 (0.68 to 1.00)	0.045
over weight	0.65 (0.51 to 0.82)	<0.001	0.64 (0.50 to 0.82)	<0.001
Incomplete information*	-	-	-	-
Presence/history of oral candidiasis
No	1
Yes	1.21 (1.04 to 1.40)	0.012
HAART status at recruitment
Not using HAART	1
AZT-based HAART	0.98 (0.68 to 1.42)	0.917
Non-AZT based HAART	0.91 (0.78 to 1.05)	0.193
Age, years
18 to < 30	1
30 to < 35	1.11 (0.90 to 1.36)	0.341
35 to < 41	1.01 (0.83 to 1.24)	0.899
≥ 41	1.01 (0.82 to 1.24)	0.936
Presence/history of fever
No	1
Yes	1.04 (0.90 to 1.21)	0.565

^‡Multivariable model only includes the variables where data are shown because all others were lacking significance.
*Three participants with missing BMI could not be included in the log-binomial regression model because the outcome did not vary, thus n = 397 for multivariable analysis and for BMI in univariable analysis.

ISSN: 1471-2334