Returning epidemics driven by geographic inter-connectedness (study)

Now that there's another measles outbreak in the news, and an apparent epidemic of whooping cough in California, it's worth returning to this earlier post on the rise and fall -- and rise again -- of epidemic diseases. They seem to track the inequality cycle, with rising inequality going along with rising prevalence of epidemic diseases.

Public health measures like vaccination programs, and educating people how to practice better hygiene, seem to have little to do with the changes up and down over time. For many of the major epidemic diseases, you can't really see the introduction and adoption of the vaccine when you look at the historical pattern of those diseases. In many cases, there was already a precipitous decline under way when the vaccine was introduced, and it didn't appear to accelerate the decline any faster.

That's not to say that the vaccine doesn't help an individual avoid the disease, or that a higher vaccination rate wouldn't contribute to "herd immunity" and help the entire group avoid epidemics of the disease. It's just to say that those effects are clearly swamped by something else at the macro level -- something related to the status-striving and inequality cycle.

Earlier, the main factor that I pointed to was immigration, which follows the inequality cycle. Immigrants are strivers themselves, leaving behind their country in order to make more money in America. And employers become more cost-cutting in striving times, hence more anxious to import cheap foreign labor. When competitive striving begins to decline, immigration gets cut off, as during the 1920s, after a peak in competitiveness circa WWI. Immigration stayed low throughout the Great Compression, and only became turbo-charged during the 1980s, with the return of competitive striving.

I didn't see it earlier, but a similar factor is transplant-ism within a large country. That's a form of immigration at a smaller scale, across states within a nation. If you look at the proportion of a state's residents that were born in some other state, it too follows the striving and inequality cycle. Transplants became more and more common during the Gilded Age and early 20th century, then folks settled down during the Great Compression. With the return of striving, they have decided to head off for greener pastures and revived the footloose, get-rich-quick transience of the Gold Rush era.

These two levels of greater demographic inter-connectedness are far more important for the spread of epidemic diseases because they increase the effective population size of the group of individuals who could spread the disease to one another. Small, sparse populations like the Bushmen hunter-gatherers in the Kalahari Desert are not subject to endemic person-to-person diseases like measles, small pox, and so on. Large, dense populations like ancient and modern Egypt are.

In epidemiology, the "critical community size" is the number of people needed for a disease to become a regular epidemic. If the population is too small, isolated cases or even outbreaks may occur, but not self-sustaining epidemics, let alone ones that recur time after time. The critical size differs across diseases because some diseases last longer in the sick person, and some diseases are more easily transmitted, both factors allowing a smaller population to sustain an epidemic.

The critical size has to be estimated from real-world data, and it is a clear threshold below which the disease won't become a regular epidemic. It's not that vaguely defined "small" populations are the ones that will be more disease-free. They just have to be smaller than the critical size, as estimated from empirical data.

As anyone who's read Plagues and Peoples will remember, increasing global connectedness has historically brought with it the spread of ravaging epidemics. That's because the global population changes from a high number of small closed groups to a small number of large webs. Epidemics are far more likely to propagate through a large web than within a small enclave.

In fact, a research team has shown the effect of inter-connectedness among an international "meta-population" on the persistence of epidemic diseases, looking at how island nations fare compared to mainland nations. Free full article here (if you are numerate, you can read it without any background in epidemiology or ecology).

Controlling for other relevant factors, islands are about twice as likely to see the extinction of an epidemic disease. Quarantine the whole outside world, and you've got it made in the shade.

Vaccination increased the chance of eradicating a disease, but vaccination rates haven't shifted radically over the past several decades, and are unlikely to explain the recent return of epidemic diseases. Transplant-ism, immigration, and other forms of demographic connectedness have shot through the roof, however, and are a much more plausible cause of whooping cough coming back from the dead.

Their study also looks at the role of what are euphemistically called "rescue" effects -- the disease is "rescued" from extinction by being introduced from an outside source where it flourishes. Inter-connectedness brings epidemics back from the dead. When relatively isolated nations come close to eradicating a disease, such rescue effects will play a larger role in the disease's persistence in those nations.

Proportion of the population who are migrants also has a strong negative effect on eradicating a disease. So, just being an island may not be a silver bullet -- if you happen to be globally connected. Singapore will fare worse than Vanuatu.

Their unstated conclusion (stating it would get them fired for racism): if a developed country has more or less fixed an epidemic disease, it shouldn't be allowing much demographic contact with outside nations -- even those where the disease isn't that common (that would still boost the effective population size, perhaps by an order of magnitude). But especially so when those nations have the disease running rampant.

Corollary: if a region within a nation has extinguished a disease, they shouldn't allow much demographic contact with other regions. Again, even if each region is relatively disease-free, joining them all into one great big swirl would boost their effective population size by an order of magnitude or more. But especially when one of those regions is more disease-ridden than another.

Stating that corollary would not only piss off the liberals, for whom freedom of movement is paramount; it would also unnerve those conservatives who are on board with limiting immigration. You mean we aren't even supposed to allow internal geographic churning either? Well then, how am I supposed to leave my flyover region and make shitloads of money in Silicon Valley or Wall Street?

A false model unsettles one faction of the powerful; the true model disturbs them all.

We ought to bear these lessons in mind when we look at vaccination rates, and ask, "Is that rate defined all the way up to the relevant level?" California's state-wide vaccination rates mean little when they are part of much greater effective population whose web of disease includes Central America and Pacific Rim populations.

And looking just within the nation, Utah's high vaccination rates will be effectively diluted by their demographic connectedness with Colorado next door, where vaccination rates are among the lowest in America.

By the same token, a tiny handful of Jenny McCarthy wannabes in the upper class will not be responsible for a return of whooping cough. The study estimates a critical size of unvaccinated births to be around 50,000 for whooping cough to go epidemic, and a size of perhaps 5 million unvaccinated in the population as a whole. Upper-class fads cannot introduce that many unvaccinated people into the population that quickly. Immigration, transplant-ism, and frequent travel (between and within nations) can do so easily. (The corresponding estimates for measles are 500,000 unvaccinated births or around 15 million unvaccinated total population.)

It would be a great help in the fight to eradicate disease to shift the discussion away from the micro-level of vaccinating individuals, and even from the semi-micro-level of herd immunity within a school or neighborhood. Our communities are unfortunately far more broadly connected webs than that. The focus ought to be on the macro-level processes that have re-introduced once-rare diseases, and on the macro-level solutions that could send them back to the grave -- namely, sealing off our communities into smaller enclaves.

Not to the extent of everyone hunkering in our own private nuclear fallout shelter, but small enough in population density that each enclave will be below the threshold for the major epidemic diseases. We could still communicate at larger scales if we wanted to, or ship goods across enclaves. But sustained demographic inter-mixing ought to be kept low. End immigration, and curtail transplant-ism as much as possible.

Citation: Metcalf CJE, Hampson K, Tatem AJ, Grenfell BT, Bjørnstad ON (2013). Persistence in Epidemic Metapopulations: Quantifying the Rescue Effects for Measles, Mumps, Rubella and Whooping Cough. PLoS ONE 8(9):e74696. doi:10.1371/journal.pone.0074696