HIV/AIDS Skepticism

Pointing to evidence that HIV is not the necessary and sufficient cause of AIDS

Deaths from “HIV disease”: Why has the median age drifted upwards?

Posted by Henry Bauer on 2009/02/18

In “HIV, AIDS, and age: HIV/AIDS theory is wrong” [23 January 2009], I pointed to the similarity of median ages for HIV tests, for “new infections”, for AIDS diagnoses, and for deaths from “HIV disease”, arguing that this contradicts the widely accepted theory that, after a “latent period” of about 10 years, HIV causes AIDS and eventual death. Not only that: all those ages are within the range of 35-45 years, whereas the maximum risk of incurring sexually transmitted infections is at ages that are younger by a couple of decades and the maximum risk of death from any illness or disease rises at an exponential rate with increasing age from around 30. Furthermore, the age distributions of deaths are narrower than the others [No HIV “latent period”: dotting i’s and crossing t’s, 21 September 2008], the very opposite of what HIV/AIDS theory demands. That theory is simply wrong, on these several counts as well as others, for example, that “HIV” tests in the United States demonstrate that those tests do not track something infectious [The Origin, Persistence and Failings of HIV/AIDS Theory].

All the mentioned median ages are similar to one another, but all of them (and of course the age distributions) have drifted upward over the years. Defenders of the orthodoxy have claimed that the upward drift in median age of death from “HIV disease” illustrates a life-extending benefit from increasingly better treatment over the years, for example:

“The median age at death due to HIV disease increased almost linearly from 36 years in 1987 to 39 years in 1995, and to 45 years in 2005. This is a reflection of the postponement to older ages of HIV-attributable deaths that were not entirely prevented by improved treatment. The median age at death due to HIV disease varied little by racial/ethnic groups”.

That statement is an annotation to Figure 1:


This suggested explanation for the upward drift of the median age of deaths ignores that the median ages have also drifted upward (1) for first positive “HIV” tests, (2) for new diagnoses of “AIDS”, and therefore, (3) of surviving PWA:


The actual numbers for the data in Figure 2 are given in Table 1:

I don’t know why the CDC graph begins at 1987 rather than 1982, which was the first year for which CDC published deaths by age group. In any case, there is not an “almost linear” increase since 1987, as the numbers in Table 1 show. From 1982 to 1987, the increase was ~0.16 years per year (YPY).  If one chooses the interval from 1982 to ca. 1990, it was a bit bigger, ~0.2 YPY; to the mid-1990s, it was a bit bigger again, approaching 0.3 YPY; for the whole period 1982 to 2004, it was ~0.42 YPY. In other words, the rate of upward drift was increasing over the years; and it was already ~0.16 YPY before there was any antiretroviral therapy. For the pre-HAART, largely AZT period 1987 to 1993, the upward drift was ~0.37; for the HAART era (1996 on), it’s been  ~0.64 YPY.

One might them be tempted to credit AZT with ~0.21 YPY benefit (0.37 – 0.16), and HAART with an additional ~0.27 YPY (0.64 – 0.37) with respect to “deaths that were not entirely prevented by improved treatment”. But even so minimal a claim ignores the fact that the ages at which people were testing positive, and the ages at which they were being diagnosed as having AIDS, had also been drifting upward at somewhat comparable rates. The population of PWA is selected by AIDS diagnoses. As that population ages, deaths among that group will also occur at increasingly older ages (just as among the population at large, life expectancy increases with age among those surviving to any given age).

The upward drift in median age of those testing “HIV-positive” was ~0.4 YPY between 1995 and 2004, among the roughly 2 million people tested annually. That was not significantly owing to any trend to test older people, for the median age of those being tested increased by only ~0.1 YPY in that period, leaving ~0.3 YPY attributable to whatever was contributing to an upward drift in age of those testing “HIV-positive”. That might be owing to changes in the tests themselves or to changes in the nature of the tested population, or of course both.

Now, the tendency to test “HIV-positive” varies somewhat by sex and race, and so does the median age at which positive tests are most likely. As it happens, the median ages at which blacks test “HIV-positive” are about a year higher than the median ages at which whites test “HIV-positive”; and for both blacks and whites, the median age at which men test positive most often is a couple of years higher than the age at which women test positive most often:


Now, the proportions as to sex and race of those being tested changed over the years. In 1995, 1.39 million females were tested and 1.10 million males, a ratio of 1.25, 25% more females than males, whereas in 2004 it was 947,000 females and 933,000 males, a ratio of only 1.02, almost equally males and females. Since the proportion of men among those tested increased significantly, and the median age for testing positive is greater for men than for women (by about 2 years, see Table 2), the overall median age for testing positive drifted upward solely because of the changing composition of PWA in terms of men and women.

Furthermore, the proportion of blacks among those tested increased significantly between 1995 and 2004: in 2004, only 36% of those tested at public sites were white and 39% black, whereas in 1995 it had been 49% white and only 33% black. That change is consistent with and perhaps a consequence of the increasingly pervasive shibboleth that HIV/AIDS has, in the United States, become a disease of the black community, spurring concerted efforts for more comprehensive testing in that community. Since blacks test positive at median ages greater by about a year (Table 2) than the median age at which whites test positive, the changing composition of PWA in terms of race adds a further upward drift in the median age of those testing positive, and thereby of the PWA population as a whole.

The most significant change in the population of PWA, however, was one that began in 1993, when the definition of “AIDS” was broadened to include healthy, asymptomatic individuals if they were “HIV-positive” and had CD4 counts below 200. Thus the average level of health among PWA was improving since 1993, and one would expect to see the median ages of those dying among that group to drift upward at an increasing rate. That’s what the data indeed show: the rate of increase was ~0.3 YPY from 1982 to 1993, and since 1999 it’s been ~0.6 (an international re-definition of “AIDS” in 1998 makes numbers from 1997-99 less comparable to others).

So the median age of death from “HIV disease” experienced an upward drift because the median age of the population of PWA, from which those deaths are drawn, drifted upward owing to the changing composition of the population of PWA in terms of age and sex. In other words, the changing composition of the population of PWA contributed indirectly to an upward drift in median age of death.

In addition to that, however, the changing composition of the population also contributed directly to an increase in the median age of death: “HIV disease” mortality among blacks is greater than among whites, and blacks also die at older ages (Table 3).

Data for intermediate years are quite similar (before 1999, reports were not in the same format). At any rate, compare male whites with male blacks, and white females with black females: in both cases, the maximum rate of death is at an older age among blacks than among whites, and in both cases the maximum rates of death are considerably greater among blacks than among whites — by a factor of ~7 with males and ~12 with females.

Since the proportion of blacks among PWA increased over the years, and since black PWA died at a greater rate than white PWA, and at older ages, the overall median age of deaths “from HIV disease” drifted upward even more than the upward drift predicted solely by the greater proportion of blacks entering the population of PWA.

So: Does the upward drift in age of death reflect some lingering benefit from HAART, as the Centers for Disease Control and Prevention suggested?

Not at all. The changing population of PWA and the different characteristics of blacks and whites suffice to explain the upward drifts of median ages of death as well as of PWA.

6 Responses to “Deaths from “HIV disease”: Why has the median age drifted upwards?”

  1. David said

    Dr. Bauer,

    Do we have any data on the consistency of HAART consumption by patients? I am a health-care practitioner, and recently saw a patient who told me that, due to side effects, he has taken frequent breaks or drug holidays since going on the meds in ’99. In fact, he abstained from the meds for an entire year in one of his holidays since first starting to consume HAART. Unless we have a better idea of the consistency of HAART usage, how can the CDC’s conclusion be taken without a grain of salt? Of course, the mainstreamers would say that inconsistency of HAART consumption might lessen its beneficial impact while we would tend to believe that HAART’s impact would look more negative had there been no drug holidays.

    • Henry Bauer said


      I don’t believe the data exist. There have been specific studies of interrupted vs. continuous HAART, and I’ve seen conclusions pro and con. But I doubt that it would even be possible to get reliable data — there are anecdotal accounts of people pretending to take the meds, or to give them to their children, for fear of the consequences if it was known that they were not taking them or administering them. Reading between the lines of the Treatment Guidelines suggests that switching of meds is quite frequent, both because of “side” effects and because they don’t work (immunologic or virologic failure).

  2. Dr. Bauer, do you have any updates for this aspect of AIDS? Have the median ages for death, first poz test, etc. continued to increase? thanks!

    • Henry Bauer said

      David Blake Jones:
      Sorry, I don’t have more information. I stopped trying to make sense of the official numbers when the Reports began, in the late 1990s, to give estimates from computer models instead of actual reported counts (I noted that at p. 245 in The Origin, Persistence and Failings of HIV/AIDS Theory).

  3. Do you think a more fruitful avenue of critique might be to look at the rates of diagnosis and death relative to the past? After all, couldn’t the orthodoxy just say that you are looking at a relatively small number of people relative to the past, who have been diagnosed or who have died and that the vast majority of people who have tested poz ARE living longer till diagnosis and death. The argument would be that there is a small percentage of patients who do not have success with the meds. The last year of data from a search I did a couple of years ago, was for 2009. There were almost as many deaths from AIDS in that year as there were in 1999, so that makes your critique more compelling, especially if the rates of diagnosis and death have started to increase. I have started to see more obituaries in the gay press for middle aged men who have died from AIDS, but that is anecdotal.

    • Henry Bauer said

      David Blake Jones:

      What you say makes sense — in principle, But I see no way to find data that’s sufficiently reliable and conclusive to change minds, unfortunately.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s