Networks and Pathogens

NETWORKS AND PATHOGENS; social networks and infectious diseases. To some, these phrases may well seem examples of the limitless extensibility of the English language and the irrepressible verbal creativity of social scientists. After all, not too long ago we had friends; now, we have a “network.” In the past, meetings were gatherings where professionals made contacts; now, we “network,” and much “networking” occurs at conferences. Surely, the topic of networks—excepting perhaps biochemical, neural, or vascular networks—is a passing fad, with limited pedigree and few prospects . . . no, not so, for many reasons. The origins of network analysis go back more than 100 years in the social sciences. For example, in his Gemeinschaft und Gesellschaft [Community and Society], 1 originally published in 1887, Ferdinand Tönnies—helping to lay a foundation for the scientific study of society—clearly had the image of a social network in mind when he observed that “(i)n the same way as a person can be linked with another person, he can be united with many persons and these again can be connected with one another.”1(p243) He noted this in the context of discussing the great variety of relationships that can connect people. Georg Simmel, another oft-cited inspiration for network analysis, called attention to the importance of personal interconnections in his essay on “the Web of Group Affiliations” [orig. 1922]. 2 Humble beginnings, to be sure, but a start. Reading Tönnies and Simmel today is not very useful. One is struck by the time it took to come even this far in establishing a scientific basis for studying societies. And, it is sobering to contemplate what must still be done to develop the full range of tools needed to better understand social factors affecting the spread and control of human pathogens. However, lest one become disheartened, two points deserve consideration. Firstly, scientific studies of biological factors affecting the health of individuals did not emerge easily or quickly. A prime illustration is Claude Bernard’s lifelong effort to establish a sound scientific basis for experimental medicine in the 19th century. 3,4 Secondly, as the on-going pandemic of HIV and AIDS has repeatedly reminded us, the health of populations cannot be protected if the only science ever pursued is that related to the physiology of hosts, the ecology of vectors, the molecular biology of pathogens, and the efficacy of pharmaceuticals. Without question, there is an imperative need for medical science to develop better ways to preserve and restore the health of individuals, and certainly the areas just mentioned (and many other related areas) play an irreplaceable role. But, there is also a critical need to understand social factors that increase or interrupt the spread of pathogens. Network analysis can provide powerful tools and a means to more effective disease control. Here, the focus is on social networks rather than personal networks. Simply defined, a personal (“egocentric”) network consists of a focal person (“ego”), the persons linked to him or her, and the social relationships that connect them. In any population there are as many personal networks as persons (excepting those completely without social ties, “isolates”). A social network is the whole set of persons in a particular population, together with the links (of interest) connecting them. 5 The foundations of modern network analysis were laid in the 1930s with the emergence of sociometry. The study of Lundberg and Steele 6 is an example of an early attempt to visualize and quantify social network phenomena. Over the course of the following two decades tools for the analysis of sociometric (social) networks slowly developed and became more scientific. After a lull, a number of contributions directed attention back to networks of interpersonal interaction. These included the mathematical/empirical work of Coleman and colleagues, 7,8 that of Harary et al on mathematical graph theory in relation to social structures, 9 Mitchell’s collection of anthropological studies of networks in Africa, 10 and Frank’s work on statistical inference from network samples. 11 These contributions stimulated renewed interest in social networks. One result was a series of meetings from the mid-1970s that brought together mathematicians, social scientists, and statisticians, many of whom were previously working largely in isolation. Other results were the establishment of the international journal Social Networks and, in the mid-1980s as (personal) computers became affordable, the first comprehensive software package for network analyses, originally developed by Freeman. 12 A next step was recognizing that not only information and influence flowed through social networks, but also some human pathogens. In other times such an observation might well have passed unremarked but HIV/AIDS—the urgent need to find out how it was spreading—ensured that the study of social networks and infectious diseases gained a foothold. As implied, most initial work focused on HIV/AIDS. 13–28 However, soon it was more widely appreciated that the network paradigm was useful for understanding the spread of other pathogens. 29–41 Concurrently, network concepts and methods became more accessible. 42–44 The article by Wylie and Jolly 45 in this issue is a good example of how this approach can help to extend the frontiers of knowledge, in this instance leading to better understanding of the spread of sex transmitted diseases (STDs). In this pathbreaking study the authors looked at STDs (chlamydia and gonorrhea) in the whole of Manitoba, identified a large number of cases and contacts (>4500) during the period considered, and found a large number of connected components (∼1500), the largest consisting of 82 persons (i.e., all 82 were connected either directly or indirectly through others). The study was able to obtain a high level of cooperation and appears meticulously thorough. Their data reinforce the importance of core groups (defined in network terms), of network structure (with their initial focus being on differences between radial and linear patterns), and the role of ethnicity (aboriginal vs. nonaboriginal) in STD transmission. In addition, the relevance of geographical bridges between different parts of the province has been highlighted. The research reported, as with much good science, is work in progress. Nevertheless, it is a major step forward; it lays a firm foundation for future research in the social/sexual “laboratory” of Manitoba and elsewhere. Considering this contribution in broader perspective, a number of observations merit mention. Firstly, those responsible for the public health of other nations would do well to examine this cutting edge research in Canada and ask if they are doing as much to protect the health of their populations. Secondly, the Wylie/Jolly study was made possible by a very fruitful collaboration between public health professionals and university researchers and is a noteworthy example of the benefits of such cooperation. Accordingly, there is the general question of the responsibility of public health departments to embrace collaborative research that can lead to improved disease control. Public health professionals at the forefront actively seek collaboration with university researchers. Enlightened health departments eagerly seize such opportunities when they appear. Yet, some departments resist, in the extreme even when collaborations would entail little if any cost to them, have Institutional Review Board approval, and involve integral medical science components. The Wylie/Jolly accomplishment shows the way forward, towards greater health department/university collaboration. Their model is worth emulating as it helps to ensure that, in the course of advancing our understanding of how pathogens spread, professional responsibilities of health departments to client populations are fully met. Thirdly, “traditional” STDs are not glamour diseases. Many are perceived inaccurately as nonthreatening, readily cured, with few adverse consequences. The high costs of STDs, in personal terms and in terms of public expenditures, need not be reenumerated here. 46 Importantly, moreover, it cannot be stressed too strongly that tools developed in studying outbreaks of STDs become available for disease outbreaks with high case fatality rates, including emerging and reemerging infections. Efforts to improve STD control have much wider relevance: new tools for understanding and controlling STD transmission take their place in the toolkit to deal with the serious infectious disease threats that will emerge in the years ahead. Fourthly, network studies of infectious diseases underscore the ongoing need for systematic research on the human societies through which pathogens spread. The potential usefulness of basic knowledge about society can be illustrated with little difficulty. One well-known example is the limited—and outdated—knowledge of sexual practices at the start of the HIV/AIDS pandemic. Initially, there simply was not enough solid, up-to-date information about the social behaviors and sexual relationships in at-risk populations to allow the kind of scientifically informed response most likely to reduce spread. Another example comes from Africa. That is, in their social research on networks in African societies in the post-World War II period British anthropologists identified many of the factors later shown to potentiate HIV transmission. They studied the rural-urban migration and travel that accompanied the transition to postcolonial societies, the conflict and normlessness of towns newly populated by disparate tribal groups, the associated adultery and multiple sexual partners, and so on. 10,47 Had it been possible to expand this line of research, and had the link with human pathogen transmission been pursued sooner, the essential elements of an effective response to HIV/AIDS in Africa could have been available much earlier. The pathogens of tomorrow may be transmitted from person to person through sexual contact, or in other ways. Hence, it is vital to maintain current knowledge about what people do, with whom they interact and in what ways, where and how they live, how they are employed and what their work entails, where they go and how they get there, what kinds of their activities link them into potential transmission cycles, etc. Just as we need basic research—work not tied to any contemporary threat—on pathogens, on vectors, on host immune systems, to name a few, so also do we need basic research on individual behaviors, on personal relationships, on social networks, and on societies. Pragmatically, there are limits to the amount of such research that will be funded in even the wealthiest nations. Hence, the dual benefits of collaborative research—bringing together epidemiologists, medical scientists, and social scientists (among others)—should not be overlooked. On the one hand, this can lead to new tools for disease control. On the other hand, perhaps especially in the case of sexually transmitted pathogens, collaboration can result in new insights about otherwise hidden facets of modern societies. Finally, as, clearly, the Wylie/Jolly study would not have been possible without health department cooperation, Manitoba Health and its public health professionals deserve special commendation for their substantial role in this Canadian contribution to the advance of knowledge.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.