Smoking and Mortality in HIV Infection
Smoking and Mortality in HIV Infection
Crothers K, Goulet JL, Rodriguez-Barrada MC, et al
AIDS Educ Prev. 2009;21(3 Suppl):40-53
Smoking is a major cause of death in HIV-negative persons and is highly prevalent in people with HIV infection. The current study addresses the effect of smoking on mortality in people with HIV infection in the era of highly active antiretroviral therapy (HAART).
Methods. The investigators used the Veterans Aging Cohort Study, which is an observational cohort study of 1034 veterans with HIV infection. These veterans were "block-matched" for age and site of care with 739 HIV-negative veterans. Smoking history was by self-administered questionnaire. Smokers were defined as "current" if they used tobacco during the past 4 weeks and as "former" if they quit smoking more than 4 weeks earlier. To estimate pack-years, the questionnaire asked respondents the number of cigarettes per day that they smoked and the duration of smoking. Cigarettes per day was truncated at more than 4 packs per day (claimed by 10 participants) and smoking duration that exceeded the patient's age.
Results. The frequency of current smokers among participants with HIV infection was 46% compared with 35% of participants who were HIV negative (P ≤ .001). The median pack-years of smoking for participants with HIV infection was 15. Mortality data were analyzed separately for participants with HIV infection (n = 734) and those without HIV infection (n = 739). Comparative mortality rates in these 2 categories for participants who never smoked, former smokers, and current smokers are shown in the Table.
Table. Adjusted Incidence Rate of Mortality
Adjusted for baseline CD4 count, HIV viral load, and use of highly active antiretroviral therapy.
P ≤ .05 compared with HIV-positive participants who never smoked.
The results showed a striking correlation between smoking and mortality that was shown to be significantly higher in participants with HIV infection than in those who were HIV seronegative. The data for current smokers showed a 131% increase in mortality, compared with 32% among those who never smoked. A significant difference in mortality in participants without HIV infection was observed, but the increased risk from smoking was much lower than that from HIV infection. Of interest, neither group showed a strong correlation with pack-years smoked.
Conclusion. The investigators concluded that current cigarette smoking was associated with a substantial increase in mortality for people with HIV infection and that even low levels of exposure had a substantial effect on mortality.
The article states the obvious, but the point certainly bears emphasis. The significance of the difference between smokers and nonsmokers was extraordinary. Although "noise" from confounding variables was clear, it was not considered in the analysis. Of note, the data were adjusted for CD4 count, HIV viral load, and use of HAART.
Abstract
Impact of Cigarette Smoking on Mortality in HIV-Positive and HIV-Negative Veterans
Crothers K, Goulet JL, Rodriguez-Barrada MC, et al
AIDS Educ Prev. 2009;21(3 Suppl):40-53
Smoking is a major cause of death in HIV-negative persons and is highly prevalent in people with HIV infection. The current study addresses the effect of smoking on mortality in people with HIV infection in the era of highly active antiretroviral therapy (HAART).
Article Summary
Methods. The investigators used the Veterans Aging Cohort Study, which is an observational cohort study of 1034 veterans with HIV infection. These veterans were "block-matched" for age and site of care with 739 HIV-negative veterans. Smoking history was by self-administered questionnaire. Smokers were defined as "current" if they used tobacco during the past 4 weeks and as "former" if they quit smoking more than 4 weeks earlier. To estimate pack-years, the questionnaire asked respondents the number of cigarettes per day that they smoked and the duration of smoking. Cigarettes per day was truncated at more than 4 packs per day (claimed by 10 participants) and smoking duration that exceeded the patient's age.
Results. The frequency of current smokers among participants with HIV infection was 46% compared with 35% of participants who were HIV negative (P ≤ .001). The median pack-years of smoking for participants with HIV infection was 15. Mortality data were analyzed separately for participants with HIV infection (n = 734) and those without HIV infection (n = 739). Comparative mortality rates in these 2 categories for participants who never smoked, former smokers, and current smokers are shown in the Table.
Table. Adjusted Incidence Rate of Mortality
| HIV Positive (n = 734) |
HIV Negative (n = 739) |
|
| Never smoked | Reference | Reference |
| Former smoker | 1.29 (0.81-2.04) | 0.83 (0.43-1.61) |
| Current smoker | 2.31 (1.53-3.49) | 1.32 (0.67-2.61) |
| Pack-years <20 >20 |
1.82 (1.20-2.76) 1.87 (1.21-2.89) |
1.03 (0.53-2.00) 0.99 (0.51-1.91) |
Adjusted for baseline CD4 count, HIV viral load, and use of highly active antiretroviral therapy.
P ≤ .05 compared with HIV-positive participants who never smoked.
The results showed a striking correlation between smoking and mortality that was shown to be significantly higher in participants with HIV infection than in those who were HIV seronegative. The data for current smokers showed a 131% increase in mortality, compared with 32% among those who never smoked. A significant difference in mortality in participants without HIV infection was observed, but the increased risk from smoking was much lower than that from HIV infection. Of interest, neither group showed a strong correlation with pack-years smoked.
Conclusion. The investigators concluded that current cigarette smoking was associated with a substantial increase in mortality for people with HIV infection and that even low levels of exposure had a substantial effect on mortality.
Viewpoint
The article states the obvious, but the point certainly bears emphasis. The significance of the difference between smokers and nonsmokers was extraordinary. Although "noise" from confounding variables was clear, it was not considered in the analysis. Of note, the data were adjusted for CD4 count, HIV viral load, and use of HAART.
Abstract
Source...