Although we were not able to assess event rates beyond 3 years of having ceased
smoking, our results LY2109761 in vivo show that the clinical benefits observed in the general HIV-negative population are also seen in HIV-positive patients. Unlike the CVD endpoints, we did not observe a decrease in the mortality rates for patients who stopped smoking during follow-up. Even when we restricted our analysis to those >50 years old, a population at increased risk of the detrimental effects of smoking, mortality rates did not decrease with passing years of having stopped smoking. These findings are in contrast to those reported in the literature for the general population both for all-cause mortality and for specific causes of death [24,26,27]. One possible explanation for our findings is that some patients who stopped smoking following diagnosis of a serious illness such as lung cancer may have stopped smoking too late to benefit from it. We do not collect reasons for stopping smoking, and also have only just begun collecting information on other serious non-AIDS-related endpoints such as non-AIDS-related malignancies, and so were not able to attempt to adjust for this bias. We were, however, able to summarize data on causes of death to assess this. We found that a larger proportion of previous smokers and those who stopped smoking during follow-up died from non-AIDS-related malignancies
compared with never smokers, while a larger proportion of never smokers died from HIV/AIDS compared with all the smoking groups. This lends some support to the notion see more that some patients who died probably stopped smoking at too late a stage of their illness to benefit selleck inhibitor from stopping. It is also notable that most reported causes of death were not directly associated with smoking, suggesting that we might have missed a reduction in smoking-related mortality because of competing risks. We did assess reductions in CVD-related mortality only, but did not see any clear reduction, perhaps because of the relatively small numbers of CVD-related deaths. It may
be that this issue will become clearer with further follow-up in D:A:D, and the recent inclusion of data on serious non-AIDS-related endpoints. The question of whether the rates of CVD in HIV-positive patients return to levels observed in nonsmokers after 5 or more years, as observed in the general population, remains unanswered. Indeed, although we observed a decrease in CVD on stopping smoking that is qualitatively similar to the trends seen in HIV-negative populations, making exact quantitative comparisons is not possible. There are a number of limitations to our analyses. First, we do not collect start or stop dates for smoking, and also do not collect data on smoking exposure such as pack-years. We were therefore only able to determine the time since stopping smoking with any accuracy for patients who reported stopping smoking during follow-up.