(A) The relationship history of an example couple. Male (M, upper) and female (F, lower) branches begin at each partner’s sexual debut and then join together into a single thick gray line when they form a couple. Male and female partners are at risk of transmission prior to couple formation at a rate equal to the product of a transmission coefficient (βM,pre and βF,pre) and the time-varying population prevalence in the opposite gender (PF and PM). Transmission after the couple has formed from extra-couple partners is similarly dependent on the population prevalence. Infected individuals infect their stable partner at a rate equal to the product of a chronic phase transmission rate (βM,within or βF,within) and the relative hazard of their current disease phase versus the chronic phase (not shown). Once infected, individuals are given Weibull distributed survival times [33,34] (not shown). (B) A simulated time series of infection, AIDS mortality, and censorship histories for ten couples. Small arrows indicate longitudinal observations of each couple, up to five times at 10-mo intervals if they have already formed at the start of observation, if they are not censored due to loss to follow-up or couple dissolution, and if both partners remain alive. These observations are then used to create a retrospective cohort (Fig. 4).