Identifying clinical and social factors influencing changes in CD4+ count in HIV infected adults in Saskatoon, Canada
Context and Rationale: The incidence of HIV in Saskatchewan is the highest in Canada and the epidemiology in this province is unique. Clinically, rapid decline in CD4+ cell count and rapid progression to AIDS has been reported. The high proportion of co-morbidities present in this population could offer some explanation for this accelerated progression. The purpose of this study was to assess the impact of selected cofactors on the rate of CD4+ count change as well as to identify factors associated with an increased risk of rapid CD4+ count decline among HIV-infected adults in Saskatoon, Saskatchewan. Methods: This is a retrospective longitudinal cohort study from a medical chart review at the Positive Living Program and the Westside Community Clinic in Saskatoon. Study inclusion criteria included HIV diagnosis between January 1st, 2003 and November 30th, 2011 and at least 18 years of age at time of diagnosis. Multiple statistical analyses were used to assess the influence of selected cofactors on changes in CD4+ count over time, including linear regression, mixed effects models and logistic regression. Results: In total 457 patients were eligible for inclusion in the study. The mean follow-up time was 46.3 (SD = ± 26.8) months. When CD4+ slope was estimated by linear regression patients who were not prescribed ARV during the follow-up period had a significantly higher CD4+ at diagnosis as compared to patients who were prescribed ARV (p<0.0001). Hepatitis C infection (HCV) (p=0.0079), a history of injection drug use (IDU) (p=0.0002), a record of incarceration (p=0.016) and not having been prescribed antiretroviral therapy (ARV) during the follow-up period (p=0.0004) were associated with a significantly more rapid decline in CD4+ count. Due to high association between First Nation or Métis ethnicity, HCV-coinfection and IDU three separate multivariate mixed effects models were built. In the model including First Nations or Métis ethnicity, First Nations or Métis ethnicity (p=0.028), receipt of social assistance (p=0.011) and age at diagnosis (p=0.0011) were significantly associated with CD4+ count. Receipt of ARV over time was significantly associated with a rise in CD4+ count (p=0.0089). In another model including HCV co-infection, HCV co-infection (p=0.0048) and age at diagnosis (p=0.039) were significantly associated with CD4+ count. Receipt of ARV over time (p=0.0004) was again associated with an increase in CD4+ count. Finally, in the third model, which included history of IDU, history of IDU (p=0.047), receipt of social assistance (p=0.042) and age at diagnosis (p=0.018) were significantly associated with CD4+ count. Receipt of ARV over time (p=0.001) was associated with an increase in CD4+ count. Model 1 had the lowest AIC value. Only CD4+ counts recorded while patients were not receiving ARV were included in the logistic regression analysis. Two subpopulations were analyzed in these models, patients with an initial CD4+ of ≥ 500 cells/mm3 and patients with an initial CD4+ count of ≥ 300 cells/mm3. CD4+ count slopes were estimated using both linear regression and mixed effects models. In the logistic regression models of subgroups which consisted of patients with the 25% steepest and 25% shallowest CD4+ count slopes, ARV was consistently significantly associated with experiencing rapid CD4+ count decline. Conclusion: HCV co-infection, a history of IDU, a record of incarceration and not receiving ARV during the follow-up period were associated with a significantly more rapid CD4+ count decline. First Nations ethnicity, HCV and IDU are highly correlated, therefore the effects of each of these variables on CD4+ count are likely not independent. Overall among all three multivariate mixed effects models, First Nation or Métis ethnicity, HCV co-infection, a history of IDU, receipt of social assistance, and age at diagnosis were associated with lower CD4+ cell count, whereas ARV treatment was associated with increasing counts. In the logistic regression models receipt of ARV was associated with rapid CD4+ count decline. Individuals exhibiting factors associated with rapid CD4+ count decline or lower CD4+ counts over the follow-up period could benefit from more frequent follow-up by clinicians and earlier initiation of ARV. Increased attention and resources focused on this population are needed to prevent disease progression.
DegreeMaster of Science (M.Sc.)
DepartmentCommunity Health and Epidemiology
ProgramCommunity and Population Health Science
SupervisorLim, June (Hyun-Ja); Skinner, Stuart
CommitteeJanzen, Bonnie; Reeder, Bruce
Copyright DateAugust 2012