Thread by @svscarpino

The intensity of #COVID19 epidemics is heavily influenced by population structure. Our new paper analyzing high-resolution case, population, & mobility data from China and Italy is out today in @NatureMedicine. Co-led w/ @MOUGK & @EvolveDotZoo. 1/15 nature.com/articles/s4159…

The work brought together an international team of collaborators: B. Rader, @AnjalikaNande, @alison_l_hill, B. Adlam, @RCReinerJr, @davidmpigott, @B_Gutierrez_G, A. Zarebski, @munikShrestha, @johnbrownstein, @marciacastrorj, @chris0dye, H. Tian, @EvolveDotZoo, & @MOUGK 2/15

Using case data from the "Open COVID-19 Data Working Group" (github.com/beoutbreakprep…), paired with high-resolution population and mobility data, we showed that epidemics are sharper in lower-density areas and broader and longer in big cities. 3/15

A similar relationship was shown for influenza by Dalziel et al. (science.sciencemag.org/content/362/64…) and was predicted theoretically by @duncanjwatts, @robymuhamad, DC Medina, & @peterdodds in 2005 (pnas.org/content/102/32…) 4/15

One of my favorite papers of all time (Sattenspiel & Herring 98) couples data from fur traders during the 1918 flu pandemic in Canada to meta-population models to show a similar effect of hierarchical structure on the shape of epidemic curves. jstor.org/stable/41465622 5/15

The work builds off of classical theory in ecology about crowding, which was developed by Lloyd in the 1960s (See

https://twitter.com/Ben_Sheldon_EGI/status/1313103441972154369?s=20

). &, if you're interested in learning more, I'd suggest this review by M. Wade, @clynfitzpatrick, & @curt_lively. 6/15 besjournals.onlinelibrary.wiley.com/doi/full/10.11…

In addition, there's some foundational work on synchrony and meta-population structure by Keeling et al. who showed, among many other things, that coupling between populations can lead to global persistence of infectious diseases. 7/15 sciencedirect.com/science/articl…

And in the network science literature by Barthélemy et al., who studied epidemics in heterogeneous networks. sciencedirect.com/science/articl… 8/15

Back to our study, using a hierarchical network model and stochastic simulations (led by @alison_l_hill) we showed how #COVID19 will spread through different kinds of communities and how the effect of interventions *depends* on the population structure. 9/15

Similar results were found by @vcolizza & @alexvespi 2008 (sciencedirect.com/science/articl…) and also @sanmelons et al. in 2011, where they demonstrated the complex dynamics of network structure and behavior on epidemics. 10/15 nature.com/articles/srep0…

Finally, we coupled our model from China and Italy with global mobility and population data to forecast the epidemic intensity of #COVID19 in cities across the planet. 11/15

If you'd like to learn more about how our work may explain observed differences between urban/rural areas and between cities like
São Paulo and Manaus, read the incredible pieces by the @UniofOxford & @Northeastern News teams ox.ac.uk/news/2020-10-0… news.northeastern.edu/2020/10/05/why… 12/15

There's tons of great work coming out on #COVID19 and population structure, including: pnas.org/content/117/39… and pnas.org/content/early/…. Now is a very important time for this kind of research! 13/15

We are in-debt to the Open #COVID19 Data Working Group. And there's some really exciting news coming about this project. Follow @globaldothealth for updates. 14/15

And we were so fortunate to have flexible and dedicated funding support from @Googleorg, @EU_Commission H2020 program, brancoweissfellowship.org, @oxmartinschool, @Northeastern, @NUnetsi, the Beijing Science and Technology Planning Project, and the @NIH. 15/15