Sexual networks and disease transmission liljeros 2003
Reviews implications of social network analysis on epidemiology of sexually transmitted diseases.
The core of the review is that STDs do not exhibit the weakly homogenous mixing which underlies the standard epidemiological model.
Traditional approach: divide into subpopulations based on number of different partners. Assortive mixing– swingers have many contacts with other swingers. This generates a faster spread and smaller size epidemic. Dissasortive interaction has swingers mixing it up with virgins, which leads to slower initial spread and a larger epidemic. Most studies find that the assortive mixing model best fits reality (refs 17 and 18). Extending these models to include all the dimensions of human interaction soon results in a model with too many dimensions to be analytically solvable (ref 30, 31).
Claims first SNA paper was by Jacob Moreno, who drew sociograms of relationships (Moreno and Jennings, “Statistics of social configurations”, Sociometry 1938). HIV is what convinced the world that the network perspective is crucial (A.S. Klovdahl, Networks and pathogens, Sex. Transm. Dis. 2001).
The Resina data is similar to other patterns– a large network but few large components (J.L. Wylie, A. Jolly, Patterns of chlamydia and gonorrhea infection in sexual networks in Manitoba, Canada, Sex. Transm. Dis. 2001)
And then the review peters out. The only real finding is that diseases in scale-free networks are always epidemics, because the variance of the degree distribution is infinite.