Today’s HIV prevention, care, and treatment strategies are based largely on the science and insights of biomedicine and epidemiology, two disciplines that have traditionally emphasised biological interventions and individual behaviour change over measures that address social or structural sources of risk. Most research has focused on the biological co-factors that affect transmission dynamics, such as the presence of concomitant sexually transmitted infections (STIs), the level of viral load in the bloodstream, or the use of condoms or other prevention methods that reduce the likelihood of transmission. But what about non-biological factors that influence behaviour and the likelihood of transmission, such as alcohol use immediately prior to sex, internalised stigma, economic and consumer pressures that encourage transactional sex, or exposure to violence and/or the impact of rigid gender norms? These factors also affect HIV transmission, but they operate earlier in the causal chain through more varied and complex pathways.
Consistent with its roots in biomedicine, HIV prevention science has traditionally emphasised expanding access to biomedical prevention tools, such as STI treatment, medical male circumcision (MMC), treatment as prevention TasP), and condoms. Such biomedical interventions are important, but need to be complemented by responses that address the structural drivers of HIV vulnerability. Indeed, if public health and HIV prevention were more grounded in the social sciences—sociology, economics, cultural studies, and social psychology—it is likely that today’s HIV programmes would look vastly different . They would place greater emphasis on context and on the social, economic, and political forces that condition people’s behaviour. They would recognise the “messiness” of real life, and acknowledge that there is seldom a single pathway that universally predicts the association 1 between distal factors—such as migration for work—and HIV acquisition. Rather, diverse pathways may operate for different individuals in different settings.