The impact of human mobility on the spread of HIV is often recognized in the medical literature (http://sti.bmj.com/content/78/suppl_1/i91.abstract?sid=2b7658a8-4f84-4d8a-b2bc-d5104c523180). Does it follow that the existence and development of transport infrastructure may have had its part in the history of the epidemic?  Can we go further, and read the whole history of an epidemic like HIV in terms of “spatial accessibility”?

Tatem & Salemi (AIDS journal) adopt a geographically-based approach, setting out to demonstrate that “the HIV/AIDS pandemic worldwide is a travel story whose episodes can be traced by molecular tools and epidemiology” (http://journals.lww.com/aidsonline/Fulltext/2012/11280/Spatial_accessibility_and_the_spread_of_HIV_1.10.aspx).  The evidence on the basis of which this story is reconstructed – the medium in which we find it written – consists in the data of HIV-1 subtypes and recombinants.  “The distribution of HIV-1 subtypes in a population”, state Mumtaz & Raddad in a study of the HIV pandemic in the Middle East, “tracks the spread and evolution of the epidemic” (http://sti.bmj.com/content/87/2/101.full?sid=2b7658a8-4f84-4d8a-b2bc-d5104c523180). Tatem & Salemi go a step further than this, interpreting the story of HIV-1 subtype distribution, as represented by 72 locations across sub-Saharan Africa (SSA), in reference to a sophisticated measure of spatial accessibility (“surface friction”) that takes into account surface cover, transport network, gradient etc., in order to reflect the ease of human travel across a landscape.

The results are presented graphically in the map on p.2353.  The 72 locations clearly fall, by and large, into four or so regions (broadly: west Africa, N. Ethiopia, east African, southern Africa), each of which is characterized by a relative homogeneity in the distribution of subtypes within it, and by a relative heterogeneity in relation to the distribution of subtypes characteristic of other regions.  These “regions” also show up on the map as continuous areas of high connectivity separated from each other by areas of low, or lower, accessibility.  Interestingly, there are four or five locations in central Africa (characterized by relatively low accessibility) which fail to show the pattern of strong dominance by one or two subtypes that we find in each of the regions of high accessibility.

There appears, therefore, to be a pattern of subtype dominance within continuous areas of high accessibility.  How does this accord with what we know about the epidemiology of HIV/AIDS?  The relative subtype homogeneity across accessible regions seems to reflect the swift diffusion of the infection across those regions.  Conversely, the greater diversity of subtypes attested in low-accessibility areas is evidence of slow spread.  The story, therefore, has two phases.  The first, during the earlier part of the twentieth century, takes place in central African subsequent to the infection crossing the species barrier.  During this period there is little diffusion, owing to poor connectivity, and the diversification of HIV-1 into its many subtypes.  The second phase, during the latter part of the century, sees the seeding of particular subtypes in different religions of high accessibility and their explosive growth thanks to good connectivity within those regions.  Subtypes A and D arrive in eastern Africa in the 1950s and 1960s respectively, whereas it is subtype C which starts the epidemic in southern Africa around 1970s, and travels to Ethiopia around 1982.

Apparently, recent analyses of the distribution of Malaria resistance markers show a similar spatial pattern to the one we see here.  The authors conclude that a comprehensive understanding of accessibility, travel and mobility in resource-poor settings could provide a valuable resource for the strategic planning of disease control.  Certainly, their study demonstrates the value of a multi-disciplinary approach to public health issues.