6
$\begingroup$

Diseases in nature tend to naturally become less virulent over time, in part, because the disease vector changes and in part because the host population develops immunity to it.

How possible is it, in theory, to accurately predict the time period until immunity to the disease or its decreased virulence causes it to be an unimportant threat to almost all human populations and how accurate could those prediction be, in theory?

Is there an effective minimum or maximum period of time over which this happens in a modern-like population?

For example, suppose that someone was designing a bioweapon in the context of a world like Earth in the near future. Could they design a disease to become obsolete after a fairly precisely measured period of time that was less than a human lifetime but more than a few months?

Improve this question
$\endgroup$
5
  • 1
    $\begingroup$ Fascinating idea. My first conundurm is that it can be hard to even measure the reproductive rate of some viruses beacuse their reproduction is complicated in real life societies, but I wonder if something can be done to work around that. $\endgroup$
    – Cort Ammon
    Apr 30, 2018 at 18:08
  • $\begingroup$ What kind of virulence would they be originally planning for? What percentage would they expect to catch it and what percentage of that would die? How easily will an immunization or cure be to role out? $\endgroup$
    – Michael Richardson
    Apr 30, 2018 at 20:11
  • $\begingroup$ @Green "to what degree of precision" is part of the question. $\endgroup$
    – ohwilleke
    Apr 30, 2018 at 22:57
  • 1
    $\begingroup$ This is really a 2-body problem; you have the disease itself and its ability to mutate, and you have 7.5 Bn or more people who are all individually distinct and whose children can ALSO carry minor mutations that can have a positive or negative effect on the disease. My view is that you can answer this statistically, but not provide a definitive answer down to individual granularity. $\endgroup$
    – Tim B II
    Apr 30, 2018 at 23:22
  • $\begingroup$ @TimBII Statistical error in the predictions does necessarily swamp your predictions at some scale and sample size. The statistics become much more meaningful with only modest increases from a very small sample size, but then, even modest reductions in statistical error in a prediction takes immense increases in sample size. You gain a huge amount between N=100 and N=1000, but not that much between N=10,000 and N=100,000. $\endgroup$
    – ohwilleke
    May 1, 2018 at 0:46

2 Answers 2

Reset to default
4
$\begingroup$

Diseases become "less virulent" by killing everybody who is not resistant to them.

When Europeans came to Americas, the diseases that they carried wiped out a good portion of Native American population:
https://en.wikipedia.org/wiki/Population_history_of_indigenous_peoples_of_the_Americas
https://indiancountrymedianetwork.com/history/events/american-history-myths-debunked-the-indians-werent-defeated-by-white-settlers/
Those diseases were not virulent in Europe b/c everybody who could die from them already did. But they remained virulent in unexposed populations.

You can try to predict which % of human population has the genes to be immune to your virus, and then use models to predict how the disease will spread and kill people. Interestingly, modern anti-epidemic and quarantine measures will slow the spread of the disease, but also slow the spread of immunity. But the virus will remain virulent on unexposed population, including your troops and your citizens. Or you can use a virus that your people have immunity to.

Edit

Virus that kills its host too fast is hardly a good application, it will not be able to spread. And infected areas will remain will remain dangerous even after all victims are dead.

One feasible idea is a flu-like virus that weakens most people without killing them, and lets them develop immunity as they recover. It will weaken your enemies enough for you to conquer them with conventional weapons, and everybody will be immune to it after a year. It will have to look like a cold, and have long incubation period to let it spread unnoticed. And you can immunize your own troops and citizens against it.

Improve this answer
$\endgroup$
1
  • $\begingroup$ "Diseases become "less virulent" by killing everybody who is not resistant to them." This happens, but it isn't the only thing that happens. Disease strains that burn too hot through the host are also eliminated because they kill all locally available hosts before the disease can spread. So, it happens through both disease level and host level selection. $\endgroup$
    – ohwilleke
    Apr 30, 2018 at 23:00
6
$\begingroup$

To answer your three questions in order:

  1. In theory you can predict non-virulence quite accurately but to do so you need a very large set of data spanning many years of real world study of the infected population and an ability to track the mutation of the disease accurately at close intervals.

  2. Ever disease is different, so is every population exposed to a given disease so I would expect that there are no hard and fast rules for a minimum/maximum period for a disease to become non-threatening.

  3. Given the unpredictable nature of mutation you'd be better off targeting the population you want to kill in such a way as to exclude the native biome, of the weapon makers, and thus render the pathogen safe for them at all times, barring unfortunate acts of god.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Re #1 The notion is how accurately can you predict it before it is introduced, not just how accurately can you predict its virulence next month or next year based upon trend lines. $\endgroup$
    – ohwilleke
    Apr 30, 2018 at 22:59
  • $\begingroup$ @ohwilleke Yeah you can't do that, that's what I'm saying, without actual data from the infection of the target population you can't make any accurate predictions. $\endgroup$
    – Ash
    May 1, 2018 at 11:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .