Trends in underlying causes of death in people with HIV from 1999 to 2011 (D:A:D):a multicohort collaboration

BACKGROUND: With the advent of effective antiretroviral treatment, the life expectancy for people with HIV is now approaching that seen in the general population. Consequently, the relative importance of other traditionally non-AIDS-related morbidities has increased. We investigated trends over time in all-cause mortality and for specific causes of death in people with HIV from 1999 to 2011.METHODS: Individuals from the Data collection on Adverse events of anti-HIV Drugs (D:A:D) study were followed up from March, 1999, until death, loss to follow-up, or Feb 1, 2011, whichever occurred first. The D:A:D study is a collaboration of 11 cohort studies following HIV-1-positive individuals receiving care at 212 clinics in Europe, USA, and Australia. All fatal events were centrally validated at the D:A:D coordinating centre using coding causes of death in HIV (CoDe) methodology. We calculated relative rates using Poisson regression.FINDINGS: 3909 of the 49,731 D:A:D study participants died during the 308,719 person-years of follow-up (crude incidence mortality rate, 12.7 per 1000 person-years [95% CI 12.3-13.1]). Leading underlying causes were: AIDS-related (1123 [29%] deaths), non-AIDS-defining cancers (590 [15%] deaths), liver disease (515 [13%] deaths), and cardiovascular disease (436 [11%] deaths). Rates of all-cause death per 1000 person-years decreased from 17.5 in 1999-2000 to 9.1 in 2009-11; we saw similar decreases in death rates per 1000 person-years over the same period for AIDS-related deaths (5.9 to 2.0), deaths from liver disease (2.7 to 0.9), and cardiovascular disease deaths (1.8 to 0.9). However, non-AIDS cancers increased slightly from 1.6 per 1000 person-years in 1999-2000 to 2.1 in 2009-11 (p=0.58). After adjustment for factors that changed over time, including CD4 cell count, we detected no decreases in AIDS-related death rates (relative rate for 2009-11 vs 1999-2000: 0.92 [0.70-1.22]). However, all-cause (0.72 [0.61-0.83]), liver disease (0.48 [0.32-0.74]), and cardiovascular disease (0.33 [0.20-0.53) death rates still decreased over time. The percentage of all deaths that were AIDS-related (87/256 [34%] in 1999-2000 and 141/627 [22%] in 2009-11) and liver-related (40/256 [16%] in 1999-2000 and 64/627 [10%] in 2009-11) decreased over time, whereas non-AIDS cancers increased (24/256 [9%] in 1999-2000 to 142/627 [23%] in 2009-11).INTERPRETATION: Recent reductions in rates of AIDS-related deaths are linked with continued improvement in CD4 cell count. We hypothesise that the substantially reduced rates of liver disease and cardiovascular disease deaths over time could be explained by improved use of non-HIV-specific preventive interventions. Non-AIDS cancer is now the leading non-AIDS cause and without any evidence of improvement.FUNDING: Oversight Committee for the Evaluation of Metabolic Complications of HAART, with representatives from academia, patient community, US Food and Drug Administration, European Medicines Agency and consortium of AbbVie, Boehringer Ingelheim, Bristol-Myers Squibb, Gilead Sciences, ViiV Healthcare, Merck, Pfizer, F Hoffmann-La Roche, and Janssen Pharmaceuticals.