Skip to main content

Table 2 Estimated changes per month in lipid levels, insulin resistance and Framingham risk score

From: Lipid levels, insulin resistance and cardiovascular risk over 96 weeks of antiretroviral therapy: a randomised controlled trial comparing low-dose stavudine and tenofovir

Variable

Increase per month

Regression coefficient

(95% CI)#

Increase per month2

Regression coefficient

(95% CI)#

p

Total-C (mmol/L)

 Stavudine

0.1275 (0.1035–0.1514)

− 0.0037 (− 0.0047 to − 0.0027)

 

 Tenofovir

0.0719 (0.0621–0.0818)

− 0.0020 (− 0.0025 to − 0.0016)

< 0.001

HDL-C (mmol/L)

 Stavudine

0.0721 (0.0572–0.0870)

− 0.0024 (− 0.0031 to − 0.0018)

 

 Tenofovir

0.0433 (0.0372–0.0495)

− 0.0012 (− 0.0014 to − 0.0009)

< 0.001

LDL-C (mmol/L)

 Stavudine

0.0516 (0.0324–0.0709)

− 0.0013 (− 0.0021 to − 0.0005)

 

 Tenofovir

0.0254 (0.0175–0.0334)

− 0.0007 (− 0.0010 to − 0.0004)

0.036

Log TG (mmol/L)

   

 Stavudine

0.0125 (− 0.0017–0.0268)

− 0.0001 (− 0.0007–0.0005)

 

 Tenofovir

0.0073 (0.0014–0.0132)

− 0.0003 (− 0.0006 to − 0.0001)

0.058

Glucose (mmol/L)

 Stavudine

0.0689 (0.0393–0.0980)

− 0.0021 (− 0.0033 to − 0.0009)

 

 Tenofovir

0.0679 (0.0558–0.0800)

− 0.0020 (− 0.0026 to − 0.0015)

0.991

Insulin (pmol/L)

 Stavudine

0.3941 (− 1.3102–2.0984)

0.0045 (− 0.0659–0.0749)

 

 Tenofovir

0.0042 (− 0.7001–0.7086)

0.0231 (− 0.0059–0.0521)

0.677

Log HOMA-IR

 Stavudine

0.0242 (0.0006–0.0479)

− 0.0004 (− 0.0014–0.0006)

 

 Tenofovir

0.0158 (0.0060–0.0255)

− 0.0002 (− 0.0006–0.0003)

0.756

Logit Framingham risk score*

 Stavudine

0.0100 (− 0.0050–0.0251)

− 0.0001 (− 0.0008–0.0005)

 

 Tenofovir

− 0.0047 (− 0.0109–0.0016)

0.0002 (0.0000–0.0005)

0.144

  1. CI confidence interval, HDL-C high density lipoprotein cholesterol, LDL-C low density lipoprotein cholesterol, OR odds ratio, TG triglycerides, Total-C total cholesterol
  2. *Data for South African site only
  3. #The model includes adjustment for age, gender, site, body mass index and viral load
  4. Month2, time was used as a quadratic term to take non-linearity into account. p values were estimated with a likelihood ratio test for both interactions. See “Methods” for details