Deaths Down for People With AIDS; Causes Shift From AIDS- to Age-Related Diseases

On Mar. 26, 1998, the first of a number of important observational database studies was published in the New England Journal of Medicine. This study by Palella et al.1 was the first to report a marked reduction of morbidity and mortality as the era of protease inhibitors, or HAART, began. Information derived from this large cohort of HIV-infected patients (HOPS or HIV Outpatient Study) regarding the trends in both the death rate and the causes of death among HIV-infected persons has been regularly updated.

The data is presented from the HOPS cohort, in which 5,561 participants, treated at two public, four university and two private clinics, have been followed from Jan. 1, 1996 to Dec. 31, 2002 for a median follow-up period of more than 35 months. In this most recent analysis, the rate of death, opportunistic disease (OD) and non-opportunistic diseases (NODs), determined to be the cause of death by disease category, have been evaluated. This latest analysis also looked at CD4 cell counts and time spent on antiretroviral therapy (ART).

As previously reported (as of June 2003), the death rate fell from 6.3 deaths per 100 person-years (PYs) of observation to 2.2 in 2002, stabilizing at ~2 deaths/100 PYs from 1998 on. Some of us have been concerned that the death rate has hit a plateau; however, in discussing the results with the lead author, Dr. Frank Palella, since the rate is so low, it may be difficult to demonstrate any further reductions. Similarly, the death rate due to ODs has declined from 23 deaths/100 PYs in 1996 to 6 deaths/100 PYs in 2002. In contrast, the death rate from NODs increased over the time period and depended upon the length of time the patient was on ART: 45% for patients on ART for two years to 70.2% for those on ART for seven years.

The percentage of patients on ART rose overall from 48% in 1996 to 80% in 2002. However, the particular types of ART used have not been able to be determined. The increased use of ART was associated with an increase in the mean CD4 cell count and the level within six months of death increased from 65.6 cells/mm3 in 1996 to almost 150 cells/mm3 (148.4) in 2002. However, viral loads were not reported. For the three-year period 2000-2002, the most frequent NOD causes of death were hepatic (35.6%), pulmonary (22.7%), cardiovascular (17.2%) and renal (9.8%).

The conclusions drawn from this paper are quite simple: The overall death rate and deaths due to ODs have continued to remain low over the past several years, while the rate due to NODs has increased. These changes have been associated with an increasing time on ART and have occurred at significantly higher CD4 cell counts. However, Dr. Palella commented that the "real" take-home message from the data analyzed is as follows: "If someone takes ART, they will live longer and when death occurs, it will not be due to an AIDS-related condition." Indeed, this is consistent with what many of us are seeing in our patients. As this occurs, age-related diseases (e.g., cardiovascular disease, diabetes mellitus, etc.) begin to play an ever-increasing cause of both morbidity and mortality. However, the current analysis does not provide us with the influence of age on any of the NODs. The HOPS database (in addition to other large cohorts; e.g., MACS, EuroSIDA, etc.) continues to provide us with important information regarding the natural history of HIV/AIDS. As is true for all observational databases, they do not give us "cause and effect" relationships, merely associations. However, these associations often lead to clinical trials that do attempt to evaluate the possible cause and effect relationships that are so necessary to make further advances in the management of HIV/AIDS.


  1. Declining Morbidity and Mortality Among Patients With Advanced Human Immunodeficiency Virus Infection, NEJM; 338: 853-860.


Abstract: Mortality and Morbidity in the HAART Era: Changing Causes of Death and Disease in the HIV Outpatient Study (Poster 872)
Authored by: F. J. Palella Jr., R. Baker, A. C. Moorman, J. Chmiel, K. Wood, S. D. Holmberg, The HOPS Investigators
Affiliations: Northwestern Univ., Chicago, IL; Cerner Corp., Herndon, VA; CDC, Atlanta, GA