A symbiotic SIR process

We study a symmetric two-disease SIR co-infection model on networks in which co-infected individuals recover at a rate distinct from that of single infections. The model explicitly represents all co-infection states and features absorbing recovered compartments for both diseases. Within a mean-field network approximation, we derive the basic reproduction number of the coupled system and show that invasion thresholds coincide with those of two independent SIR processes. Exploiting an exchange symmetry in the equal-transmission regime, we reduce the dynamics to a lower-dimensional invariant subsystem and analyze the impact of the co-infection recovery rate. We prove that slower recovery of co-infected individuals monotonically increases the co-infection burden and yields a lower bound on epidemic duration that grows as the co-infection recovery rate decreases. Numerical simulations further indicate that reduced co-infection recovery can increase the epidemic peak, an effect supported by a sensitivity-equation analysis. Together, these results highlight how co-infection-specific recovery dynamics can substantially alter transient epidemic behavior, even in the absence of endemic equilibria.

💡 Research Summary

The paper introduces a symmetric two‑disease SIR co‑infection model on networks, motivated by the symbiotic two‑species contact process. Each individual can be in one of nine states defined by the pair (X,Y) where X and Y denote the status (susceptible, infected, recovered) with respect to disease A and disease B, respectively. Transmission rates λ₁ and λ₂ govern the spread of the two diseases, while a common recovery rate μ applies to singly infected individuals. Crucially, individuals co‑infected with both diseases recover at a reduced rate μ̄ < μ, capturing a “symbiotic” recovery effect observed in real co‑circulating pathogens such as SARS‑CoV‑2 and influenza.

Using a heterogeneous mean‑field approximation on a network of average degree ⟨k⟩, the authors derive a system of nine ordinary differential equations. Linearization around the disease‑free equilibrium (DFE) yields the next‑generation matrix K = F V⁻¹, where F contains new infection terms and V contains transition terms. The spectral radius of K gives the basic reproduction number R₀ = max(R₀,A, R₀,B) with R₀,A = λ₁⟨k⟩/μ and R₀,B = λ₂⟨k⟩/μ. Hence, at the DFE the invasion thresholds are identical to those of two independent SIR processes; the co‑infection compartment does not affect the linear stability.

The authors then focus on the symmetric transmission regime λ₁ = λ₂ = λ. In this case the system is invariant under exchange of the two diseases, allowing a reduction from nine to six dimensions when initial conditions respect the symmetry. Defining a = (i,s) = (s,i), b = (r,s) = (s,r), d = (r,i) = (i,r), u = (s,s), c = (i,i), and v = (r,r), the dynamics evolve on an invariant subspace. The total infection mass M(t) = a(t)+c(t)+d(t) and the total number of infectious individuals J(t) = 2a+2d+c are expressed in these variables.

Two main analytical results are proved. First, the co‑infection prevalence c(t) satisfies
c′(t) = 2⟨k⟩λ a(t) M(t) − μ̄ c(t).
Applying a comparison argument (Proposition 1) shows that c(t) and its cumulative burden B(μ̄) = ∫₀^∞c(t)dt are strictly decreasing functions of μ̄. Thus slower recovery of co‑infected individuals (smaller μ̄) inevitably raises both instantaneous and cumulative co‑infection load.

Second, using the inequality c′ ≥ −μ̄c, the authors derive a lower bound on the epidemic duration. If at some time t₀ the co‑infection prevalence exceeds a threshold ε, then c(t) ≥ c(t₀) e^{−μ̄(t−t₀)} for all t ≥ t₀, implying that the time T_J(ε) at which the total infectious population falls below ε satisfies
T_J(ε) ≥ t₀ + (1/μ̄) ln