Send response to journal:
To Mr. Nicholas Bennett, As to the calculations, they were some numbers that I was forwarded a while back from correspondence with a person in SA. Quite honestly I'd have to go back in the discussion we were having at the time and obtain their primary data. That being said -- I am not sure that is even really necessary because the report we are both referring to clearly questions the anomaly. On page 63 it reads:
"Findings requiring further research The observation that the estimated HIV prevalence among children aged 2–14 years is 5.6% (CI: 3.7–7.4%) was unexpected. Once HIV prevalence was identified to be high in this group, a record review was undertaken to determine how many could have been infected through vertical transmission. The analysis was done focusing on those 2–11 years of age, as they are less likely to be sexually active. An analysis of the mother-child and father-child pairs revealed that of the 86 HIV positive children aged 2 to 14 years, 27 could be matched with a biological parent, and 20 of the parents selected in the study had an HIV test result. Of these 20, only five (25%) were HIV positive (four females and one male) and 15 (75%) were HIV negative. This raised the question of whether the biological parents of these children were alive or dead. The results found were that only seven (6.1%) of the 86 children had a biological mother who had died and a similar percentage (7%) had a biological father who had died. It remains unclear as to how these children could have been infected. ****An emerging theory that warrants further investigation is that there is unexplained HIV prevalence in children who have had no sexual exposure, or have parents with HIV negative mothers.**** In addition between 20–40% of HIV infections in African adults are associated with injections (Gisselquist et al. 2002). Given this unexplained high prevalence in children aged 2–14 years, it is necessary to test this theory in South Africa. Possible factors to be investigated include sexual abuse and unsterile needles." A couple of other interesting facts re these data. To my knowledge, the oral ELISA test that was used to derive those numbers was, at the time ( I am not sure now) not approved for those 13-years of age and under. So I think one has to question the data from that perspective too. I am not sure why saliva from a 12-year old is any different than saliva from a 14- year old when it comes to measuring HIV seroconversion, but apparently it either is, or the test was never approved in that age group -- however I still wonder why it would be any different."
You said: "If 6-7% of 75 year olds came back as HIV+ but zero percent reported a high-risk activity, I'd be very suspicious of those rates."
I find it interesting you that you would be "suspicious of the rates" for this age group with no risk factors, but seem to be prepared to accept the numbers in other age cohorts who seem to have no apparent risk factors. Granted we are talking a hypothetical situation here. Despite the fact it seems a relatively easy study to do, nobody seems to have has done it. It also seems a rather obvious study to do. If there is a 6-7% HIV positive rate in people of that age and if they have no risk factors (perhaps nuns/priests who were born in SA and have never received questionable medical treatment where bloodborne pathogens could be transmitted) then I think one would be compelled to rethink some of the generally accepted models of HIV. Indeed, even in the 2002 SA study, a rate of 6-7 % among those 55 and older is highly suspicious.
The nun/priest study would be an interesting one even in the developed world, but we would tell the people doing the lab work that the blood source was from highly sexually active drug-using gay males (for half of the group) and 75+ year-old nuns/priests for the other half, and see if there is bias in the matched groups due to source-blood knowledge.
Out of that study we would be able to determine:
1) Any lab report biase that might creep in via the lab interpretation (highly sexually active drug-using males vs. priests/nuns).
2) The prevalence rate of HIV seropositivity in 75+ nuns and priests with no risk factors for HIV. Even vertical transmission could be ruled out because these people would have been born before BA - Before AIDS.
If the background rate of HIV in the nuns/priests (or any other aged cohort with no risk factors) was statistically significant when compared to the general population, then the entire HIV = AIDS relationship falls apart.
You said: I'm interested in Mark Bartlett's views on the fact that HIV infection rates show a neat temporal relationship in South Africa"
I would have to know a lot more about the Hx of how these data were collected to make a reasoned response i.e., demographic shifts in sample groups/size, test methodology, etc. I would agree the temporal relationship looks interesting. Frankly it reminds me of the disease distribution curve the CDC made by changing the AIDS indicator diseases several times -- without that, the AIDS epidemic would have gone flat real fast too.
Another thing that I will be watching for in the next 10-20 years, is an increase of the heterosexual, SA-HIV variant into the developed world. There have been references that this is now happening. If this variant truly has a proclivity for heterosexual transmission, and spreads more rapidly as has been reported, the numbers should explode in the heterosexual population, and over time we will have an epidemic that is equally distributed amongst males and females in the developed world. Something to watch for.
Thanks again for the discussion - ours seems to be one of the more civil exchanges on this board.
Competing interests: None declared