June 2009

Impact of Cigarette Smoking on Mortality in HIV-Positive and HIV-Negative Veterans
Crothers K, et al. AIDS Educat Prev 2009;21:(3 Suppl)40-53. [PubMed Abstract]

Background: Smoking is a major cause and death in HIV-negative persons and highly prevalent in those with HIV infection. The current study addresses the impact of smoking on mortality in persons with HIV infection in the HAART era.

Methods: The authors used the Veterans Aging Covert 5 Site Study which is an observational cohort that includes 1,034 veterans with HIV infection. These were “block-matched” with 739 HIV-negative veterans matched for age and site of care. Smoking history was by self-administered questionnaire (available at http://www.vacohort.org/). Smokers were defined as “current” if they used within the last 4 weeks and were classified as “former” if they quit over 4 weeks previously. The questionnaire asked the number of cigarettes per day and the duration to estimate pack-years. It is noted that cigarettes/day was truncated at more than 4 packs per day (claimed by 10 subjects) and smoking duration that exceeded that patient’s age.

Results: The frequency of current smoking among those with HIV infection was 46% compared to 35% in those who were HIV negative (P=<0.001). The median pack-years for those with HIV infection was 15. Mortality data was separately analyzed for those with HIV infection (n=734) and those without HIV infection (n=739). The data for comparative mortality rates in these two categories for those who never smoked, former smokers, and current smokers are provided in the following Table (1).

   *Adjusted for baseline CD4 count, HIV viral load and use of HAART. 
   **P= <0.05 compared to HIV positive and never smoked.

The results showed a striking correlation between smoking and mortality which was shown to be significantly greater in those with HIV infection compared to those who were HIV seronegative. In fact, the data for current smokers showed a 131% increase in mortality compared to 32% among those who never smoked. There was also a significant difference in those without HIV infection, but here the increased risk of smoking was much less than with HIV infection. Interestingly, neither group showed a strong correlation with the pack-years smoked.

Conclusion: The authors concluded that current cigarette smoking was associated with a substantial increase in mortality for persons with HIV infection and that even low levels of exposure had a substantial impact.

Comment: The article says the obvious, but certainly bears emphasis. The significance of the difference between smokers and nonsmokers was extraordinary. There is obvious noise with confounding that was not considered in the analysis, but it should be noted that data in the Table were adjusted for CD4 cell count, HIV viral load, and use of HAART.