Recent comments by “Köpek Burun” (= “dog nose”?? “snout”? Menganito?!), about “Poison in South Africa” [26 October 2008] inevitably referred back to “HAART saves lives — but doesn’t prolong them!?” [17 September 2008]. In my response, I referred to calculations that I’ve been working on, of the age distributions of PWAs (“People living With AIDS”), and the age distribution of death rates. Since it’s so pertinent to that discussion, I need to post the calculations even though my full analysis isn’t finished yet.
Age distributions of PWAs can be calculated from the data in Table 2 of “HAART saves lives”. The (average) number of PWAs during a given year results from adding new diagnoses in that year to survivors at the end of the previous year. Those survivors can be calculated from the total number of diagnoses minus deaths up to and including that year. For calculating total numbers of PWAs for each year, that is straightforward, and the results were given in Table 3 of “HAART saves lives”. However, for the age distribution of PWAs in each year, one must take into account that survivors from a given year will be a year older, on average, in the following year. For example, some of the survivors from the age range 20-29 in 1990 will be in the range 30-39 in 1991. I made the assumption of symmetrical distribution within each age range — in other words, represented the data by histograms defined by the age ranges in which the data were reported. A number of trial calculations using more elaborate curve fitting showed that this did not make a significant difference to the results, presumably because the chief variable of interest is the median age and most of the cases fall in the middle age ranges (where the cases-vs.-age curves are steepest, there numbers of cases are so much smaller than in the middle age ranges that small errors there hardly affect the calculation of median age).
Another complication is that the age ranges for which deaths and diagnoses, respectively, are reported were not the same in the years 1993-98. For those years, the age ranges for the deaths were converted to those for the diagnoses, again using a histogram model; that this did not introduce drastic errors was verified by the fact that the re-calculated median ages remained within 1% of the initial ones.
The reason for dual death reports for 1998, for comparison with earlier and later years, was given in the notes to Table 2 in “HAART saves lives”. For purposes of comparison over the whole period 1982 to 2004, the most appropriate values for 1998 are presumably the average of those dual numbers, namely, 39.4 for median age of PWAs, 41.2 for median age of deaths, 1.8 for the interval between them, and 4.5% for the death rate. All of those fall smoothly into the progressions from 1982 to 2004.
Table I
I had begun this work to probe the effect of HAART, but I realized eventually that these death statistics speak directly to the issue of whether HIV causes AIDS, well beyond merely demonstrating that HAART doesn’t extend life. From 1982 to 2004, the death rate (last column in the Table) declined from 65% to 2.8%; yet the difference between the median age of the population of PWAs and their median age of death increased only from about 7 months (0.6 years) to about 22 months (1.8 years). That’s a stark contradiction. The median age of death in any population is the average life-span. If the median age of the existing population is within a couple of years of the life span, then the death rate must be enormous; but here the mortality in recent years is small while the life span is a mere 22 months greater than the median age of the population.
The contradiction means that the basis for classifying as “PWA” is not the same as what determines death; those who are dying are in some manner atypical within the PWA population. What typifies PWAs, though, is being HIV-positive. Therefore something other than being HIV-positive distinguishes those who are dying from those who are not dying.
At least for the most recent decade, HAART seems the obvious “missing link”. Let’s assume that those few percent of PWAs who are dying have not been getting HAART. Then PWAs who survive, who are benefiting from HAART, would be getting older, and the median age of the PWA population would be steadily increasing IN CONTRAST to the median age of those dying, which would continue to be that typical for untreated PWAs. The data show no such thing. During the HAART era, the median ages of death and of PWAs drift upwards in tandem, with no discernible change in the magnitude of the difference, 1.7-1.9 years. Most striking, the median age of surviving PWAs remains below, not above, the median age of death.
In any case, the same contradiction between median-age differences and mortality rates applies in the years 1982 to 1996. The only resolution for this conundrum is to recognize that what determines PWA status — namely, being HIV-positive — isn’t what determines death among PWAs. In other words, HIV doesn’t cause death.
*******************************
There’s another, independent, aspect of “HIV disease” deaths that speaks against HIV as a cause of death. Several times [for example, “How ‘AIDS Deaths’ and ‘HIV Infections’ vary with age — and WHY”, 15 September 2008] I’ve remarked on the peculiarity that the death rate for “HIV disease” is at a maximum roughly at ages 35-45, something like the prime years of adulthood.
All other diseases show the very opposite, death rates at their lowest among young-to-middle-aged adults and high among very young children and increasing progressively at ages beyond middle age. For instance, in “’HIV Disease’ is not an illness” [19 March 2008], Table B shows all-cause mortality lowest among young teens and increasing with age (very roughly, doubling in each higher decade); Table C shows a similar variation for cerebrovascular diseases; the “Health, United States” (HUS) reports from the National Center for Health Statistics display this type of variation with age for every type of illness. For influenza, here’s a graphical representation:
Figure I
from a poster presentation, “Death and Aging in the Time of Influenza: United States, 1960-2002” by Nobuko Mizoguchi, MPH/MPP Department of Demography, University of California at Berkeley]
“HIV disease” is entirely different, see for instance Figures 2a,b in “No HIV ‘latent period’: dotting i’s and crossing t’s” [21 September 2008] , or any of the tables for deaths from HIV disease in HUS reports, e.g. reproduced as Table D in “’HIV Disease’ is not an illness”, 19 March 2008 ; see also Table 2 in “HAART saves lives — but doesn’t prolong them!?”, 17 September 2008. “AIDS” deaths, be it in absolute numbers or in rates, are at a maximum around age 40 ± 5.
In my view, this alone already gives the lie to claims that HIV is fatally pathogenic. No matter what the origin may be of an attack on the living human organism, the tendency to succumb and die increases steadily with age. What could it be about “HIV” to allow older people to resist its ravages better than people in their prime middle years?
The data can’t be explained away speculatively as something about ages at which people most likely get “infected”, because EVERY age distribution having to do with HIV or with AIDS peaks in those same years: positive HIV tests, new AIDS diagnoses, deaths from “HIV disease”, median age of all PWAs (see Table I above). Age distributions for deaths and for positive HIV tests superpose, as illustrated graphically in “How ‘AIDS Deaths’ and ‘HIV Infections’ Vary with Age — and WHY”, 15 September 2008 , and in “No HIV ‘latent period’: dotting i’s and crossing t’s”, 21 September 2008 .
Yet another way to illustrate this is to calculate age-specific PWA-specific death rates, see Table II below. The usual way of reporting death rates (as in the HUS reports) is per 1000 or 100,000 for the population as a whole in the given age-group. But one might try to gain further insight into why HIV is so peculiar by looking at what proportion of PWAs in each age group die each year (deaths in that year in that age range divided by the number of PWAs in that age range in the same year). I haven’t yet done the calculation for every year, because the salient overall conclusion seems obvious enough:
Table II
These numbers are much more sensitive than the median ages are, to the various assumptions made in calculating age distributions of PWAs, so the variations are less smooth and only clear major differences should be regarded as reliable. The crucial point is quite clear, though: how little variation there is between the death rates in the various age ranges in any given year. In 1999, for example, about the same proportion of PWAs aged 35-44 died as among those aged ≥65 or among those of intermediate age. In the years 2002-2004, a smaller proportion those PWAs aged ≥55 died than of those younger PWAs aged 35-54.
This makes no sense, if PWA, “living with AIDS”, means suffering from a fatal illness that is only temporarily staved off by continual antiretroviral treatment. Older people should succumb more readily than younger people.
The only death statistics that show maximum rates among younger adults are accidents, homicide, suicide: what one might call lifestyle hazards, not biological health challenges. That accords with the hypothesis — for which there is much supporting evidence — that AIDS in the early 1980s was an epiphenomenon of the fast-lane lifestyle practiced by small groups of gay men; look back at Tony Lance’s essay on intestinal dysbiosis.
I offer another speculation as to a possible cause of death that would not discriminate much by age. A highly toxic chemical poison that’s likely to kill within a few years would probably kill old and young people at comparable rates. AZT and other antiretroviral drugs would fit that bill.
But I don’t want to conclude on so speculative a note. The fact that deaths from HIV or AIDS are maximum at ages 35-45 shows that those deaths are not the result of an infectious disease, or for that matter of any natural illness. The fact that the median age of the PWA population has been steadily lower (within about two years) than the median age of death among those people, while the mortality has declined enormously over two decades, proves that whatever caused the deaths is not what defines the category “PWA” — i.e., HIV doesn’t cause death, HIV doesn’t cause AIDS. A fortiori, the data show that HAART doesn’t extend life: the interval between median age of the PWA population of deaths among them has held steady at 1.7-1.9 years throughout the HAART era, and surviving PWAs are not living longer than those who die.
The only explanation that satisfies all the data is that testing HIV-positive is an artefact as regards illness or death. Testing HIV-positive is just a marker of some sort of physiological response to a variety of challenges.