Skip to main content

Table 4 Factors associated with the presence of anemia

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

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

2.24 (1.44 to 3.49)

<0.001

2.97 (1.85 to 4.77)

<0.001

BMI

    

  under weight

1

 

1

 

  normal

0.41 (0.18 to 0.93)

0.033

0.43 (0.19 to 0.99)

0.048

  over weight

0.21 (0.09 to 0.51)

0.001

0.17 (0.07 to 0.41)

<0.001

  Incomplete information*

0.82 (0.07 to 10.20)

0.876

1.69 (0.13 to 21.61)

0.686

CD4 count cells/μl

    

  <50

1

   

  50 to <200

0.93 (0.52 to 1.64)

0.795

  

  200 to <350

0.73 (0.38 to 1.42)

0.361

  

  ≥ 350

0.69 (0.20 to 2.42)

0.563

  

Age, years

    

  18 to < 30

1

   

  30 to < 35

1.18 (0.65 to 2.13)

0.588

  

  35 to < 41

0.98 (0.57 to 1.67)

0.928

  

  ≥ 41

1.20 (0.69 to 2.08)

0.516

  

HAART status at recruitment

   

  Not using HAART

1

   

  AZT-based HAART

1.69 (0.58 to 4.93)

0.336

  

  Non-AZT based HAART

1.02 (0.69 to 1.53)

0.911

  

Education level

    

  None

1

   

  Minimum Primary

1.89 (0.55 to 6.46)

0.312

  

  Minimum Secondary

2.03 (0.58 to 7.04)

0.265

  

  Tertiary

1.33 (0.34 to 5.18)

0.678

  

  Unknown†

-

-

  

Employment status

    

  Unemployed

1

   

  Employed

0.67 (0.38 to 1.19)

0.170

  

Marital status

    

  Not married

1

   

  Married

0.86 (0.56 to 1.30)

0.464

  

Presence/history of oral candidiasis

   

  No

1

   

  Yes

1.62 (1.01 to 2.57)

0.043

  

Presence/history of fever

    

  No

1

   

  Yes

1.28 (0.86 to 1.89)

0.221

  
  1. Multivariable model only includes variables for which data are shown below because all others were lacking significance.
  2. One participant with unknown education level was excluded from analysis because the outcome did not vary in this stratum. *Three subjects with missing BMI could not be not included in the log-binomial regression model because the outcome did not vary, thus n=397 for the multivariable analysis and for the BMI univariable analysis.