AIDS doomsday by 2028?

Conant essays

The MI6 fray: who are the press?
Marc Rich's influence in the media
Star Wars, AI and quantum computing

A global population cataclysm should occur no later than 2028 if the Third World pandemic is left unchecked, according to my simple but I think not simplistic calculation. An activist scientist, Andrew M. Sessler, agrees with me on the scope of the problem, though he is wary of the precise numbers (as am I, actually).

Another scientist, James B. Watson of 'double helix' fame, argues in his 2003 book, 'DNA: the secret of life,' that current methods are unlikely to stave off a global AIDS catastrophe, despite effective controls in developed countries. Watson, who heads a genetics research operation, argues that much-feared recombinant DNA technology (genetic engineering) may be the best hope for a fast, efficient counterattack on the HIV virus.

The classic Bernoulli epidemic equation well captures the spread of HIV among the sexually active part of the population, which I roughly estimate at about 1/3.

Here's what I did:

I used a plausible estimate for 1980 of 1 million people infected with HIV worldwide. I then used an estimate of 37 million infected worldwide in 2001. I neglected to figure in the 20 million who have died of AIDS between 1980 and 2001. Such a complication would not tend to lower the rate in the equation.

Another minor flaw in the equation is that we assume a constant population, when in fact it grew from 4.45 billion to 6.08 billion between 1980 and 2001. However, by using two different population figures in the two sample ratios, we implicitly account for population growth. And since the disease is spreading so rapidly in the Third World, where most of the earth's inhabitants live, the population constant is, practically speaking, irrelevant. However, various averaged population constants (in order to account for different rates) give nearly the same curves, with maximum population catastrophe about 2030 or sooner.

Calculation method i

The standard equation used was

y(t) = 1/[(1/a - 1)exp(-rt)]

Setting a --for 1980-- at 106/[4.45(109)] = 4450

and y(21) --the proportion for 2001-- at 37(106)/[6.08(109)] = 37/6800

we get r = .1574

This gives the following global infection percentages:

2002: 0.7 %

2003: 0.8 %

2004: 0.97%

2005: 1.14%


2010: 2.46%

2015: 5.26%

2020: 10.87%

2025: 19.17%


2028: 33%

Calculation method ii

I did another calculation based on UN HIV figures for South Africa in 1997 and 2002. The UN percentages are based on adult population, which is defined as persons aged 15 to 49. The UN figures for 1997 and 2002 are 12.6% in 1997 and 20.1% in 2002.

Estimating the adult population again at a third of total population, I plugged in the numbers to obtain a constant of proportionality r = -.09869. (Here I used a slightly different world HIV estimate of 40 million for the initial value.)

The following is a projection of HIV among the earth's total population based on the South Africa figures:

2005: 1%

2010: 1.6%

2015: 2.6%

2020: 4.2%

2025: 6.7%

2030: 11%

2035: 16.2%

2040: 24%

2043: 33%

This HIV scenario is perhaps conservative since overall adult population may well exceed 1/3 of total earth population. (Sorry for not taking the time to dredge these figures up, but such refinements aren't likely to change the picture much.)

The UN reports that when 4% of adults (as defined) have HIV, 35% of adult deaths are tied to AIDS, for an overall adult mortality rate of 75%. When 8% of adults are infected, a person aged 25 to 34 faces a 4/5 probability of imminent death.

That is, we may expect that population catastrophe begins at 1.33% of total population and reaches a crisis point at 2.6% of total population.

However, the UN back-projected its HIV estimate for South Africa to 4.5% of adults in 1994. The Bernoulli equation then gives r at 1.19704 for 1994 to 1997 and r at 0.19386 for 1994 to 2002. Those values would mean 33% of the world population infected by 2003 or by 2020.

The UN of course doesn't use anything so crude as this old-fashioned single variable equation, but rather uses the software program Epimodel, about which I know nothing. Yet we may infer that Epimodel predicts a sharp bulge in infections at around 5% of adults (or 1.67% of the overall population), with the rate tapering off somewhat later. Yet, the 5% mark seems to harbinger a brief but devastating surge in infections. This inference is in accord with a CIA analysis discussed below.

The industrial nations, through use of anti-retroviral drugs, condom availability, routine HIV screening, and aggressive public education will likely keep their AIDS rates low for some years to come. Yet an HIV rate for the rest of the world fast approaching 2.6% of total population points to widespread death through AIDS and famine (not enough healthy workers) and global economic chaos. HIV 'safe zones' face being overwhelmed by refugees from AIDS-blighted regions.

I doubt this forecast is unduly pessimistic. The HIV epidemic began, perhaps in the 1970s, with only a few infected persons in subsaharan Africa, where it spread, essentially unchecked. Depopulation and social breakdown of Botswana and Zimbabwe is near at hand with the same grim fate for other African nations closing fast.

At present, the HIV epidemic is taking off in south Asia and the Pacific. The Caribbean is ripe for a major outbreak.

A 2003 UN report (see also this site) shows that Eastern Europe, Russia and Central Asia are showing AIDS rates that point to a catastrophic pandemic of the type that has left 11 million African children orphaned.

Interestingly, the UN, faced with the most recent AIDS figures, downwardly revised its global population estimate for 2050 from 11 billion to 9.3 billion. This estimate appears unduly optimistic, perhaps assuming a worldwide campaign to counter the disease.

If the Group of 8 industrial nations lacks the political will to begin emergency intervention on a global scale, the world as we know it will cease to exist within a few decades.

In a September 2002 email, Andrew Sessler, a physicist with the Union of Concerned Scientists, wrote, 'Probably [your analysis] is roughly correct, but an overestimate due to self-limiting effects (like running out of healthy people in particular areas, etc.).

'Nevertheless, unless something is done (rich nations giving more than the U.S. is now giving, development of low-cost drugs, development of a vaccine, etc.), the general dimension of the problem is going to be as you say.'

Sessler suggested I search the literature for more accurate projections, which I will only do if I can do so efficiently. At any rate, the purpose of my back-of-the-envelope calculation was to get a picture of the general dimensions of the problem and not to get the most fine-tuned analysis possible.

Aggressive countermeasures, in particular education and condom availability, have worked wonders. Thailand turned its surging rate around, even without much availability of retroviral drugs, and the African nation of Senegal, not noted for sexual abstinence, has long kept its AIDS rate down using such measures.

However, entrenched ideologies may result in terrible consequences. An egregious example is China, where the communist line is one of denial that communist solutions to social problems are ineffective. The communists say, for example, that alcoholism has been pretty much eradicated. Hence, little in the way of treatment or prevention is available for those afflicted with that disorder.

Yet, alcoholism and the closely related phenomenon of drug use, including intravenous drug use, are known risk factors for HIV.

China's official attitude is to play down the significance of the HIV problem, perhaps not wishing to give any propaganda advantage to critics of communism. However, in June 2002, the UN sharply warned China that its do-nothing policy on AIDS would bring disastrous consequences to the people.

China's latest estimate that 1 million Chinese are infected still falls far short of estimates made by foreign observers.

But even accepting the official estimate but using calculation methodii, if HIV is left unchecked, an adult Chinese aged 25 to 34 will have a 4/5 probability of dying in 2033.

Calculation method iii

I was also able to find a figure of 25.5 million infected globally in 1996. Calculating the world population at 5.745 billion for that year, I plugged the numbers into the Bernoulli equation and obtained a much lower rate of infection globally.

These figures give this picture:

2005: 0.78%

2010: 1.07%

2015: 1.47%

2020: 2%

2025: 2.7%

2030: 3.7%


2060: 20.5%

However, I suspect that this estimate is much too conservative. That is, the five-year rate for 1996 to 2001 reflects the fact that the HIV epidemic was brought under control in industrial nations since AIDS was detected in 1981. However, the impact of the rate reversal in the industrial nations will be rapidly eclipsed as the disease spreads through the developing nations.

In June 2002, the CIA's National Intelligence Council predicted that the AIDS pandemic will rapidly worsen in the years ahead. Ethiopa's 5% infection ratio was cited as an indicator of a sudden surge in infections.

In news accounts, the NIC did not disclose its estimates of numbers of people anticipated to be infected in future years nor did it reveal what mathematical and epidemiological method was used to arrive at its assessment.

The fact that the estimate is being treated as classified intelligence data suggests that the NIC prediction is closer to the higher of my estimates.