pith. sign in

arxiv: 2604.21065 · v1 · submitted 2026-04-22 · 📡 eess.SY · cs.SY· math.DS· math.OC

On the dynamic behavior of the network SIRS epidemic model

Giulia Gatti , Giacomo Como This is my paper

Pith reviewed 2026-05-09 23:09 UTC · model grok-4.3

classification 📡 eess.SY cs.SYmath.DSmath.OC
keywords SIRS epidemic modelnetwork dynamicsbasic reproduction numbertranscritical bifurcationendemic equilibriumglobal stabilitydistributed algorithm
0
0 comments X

The pith

The basic reproduction number R0, as the dominant eigenvalue of a rescaled network interaction matrix, decides whether the disease dies out or reaches a stable endemic level in SIRS models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that for SIRS epidemic models on any connected deterministic network with possibly different recovery and loss-of-immunity rates across nodes, one scalar parameter fully governs the long-run outcome. This parameter R0 equals the largest eigenvalue of the contact matrix after each row is divided by the node's recovery rate. When R0 stays at or below 1 the infection is driven to zero from any starting condition. When R0 exceeds 1 the infection-free state loses stability and a unique positive endemic equilibrium appears that attracts all trajectories. The same analysis yields monotonicity relations for how the endemic levels respond to changes in contacts or rates and supplies a provably convergent distributed algorithm to compute them.

Core claim

For connected but otherwise general interaction patterns and heterogeneous recovery and loss-of-immunity rates, we identify a fundamental parameter R0 (the basic reproduction number), which fully characterizes the qualitative dynamic behavior of the system. This parameter is the dominant eigenvalue of a rescaled version of the interaction matrix, whose rows are normalized by the corresponding recovery rates. We prove that a transcritical bifurcation occurs as R0 crosses the threshold value 1. Specifically, we show that, if R0 does not exceed 1, then the disease-free equilibrium is globally asymptotically stable, whereas, if R0 is larger than 1, then the disease-free equilibrium is unstable,

What carries the argument

The basic reproduction number R0 defined as the dominant eigenvalue of the interaction matrix with each row scaled by the inverse of the corresponding recovery rate; it serves as the bifurcation parameter that switches the system between global extinction and a unique stable endemic state.

If this is right

  • Global convergence to the disease-free equilibrium occurs for every initial condition whenever R0 is at most 1.
  • A unique endemic equilibrium exists and is asymptotically stable whenever R0 exceeds 1.
  • The components of the endemic equilibrium increase monotonically with each entry of the interaction matrix and decrease monotonically with each recovery rate and loss-of-immunity rate.
  • A simple distributed iteration converges to the endemic equilibrium from any positive initial guess.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Interventions that lower the dominant eigenvalue of the rescaled matrix below one are guaranteed to eradicate the disease regardless of starting conditions.
  • The threshold result supplies a clear design criterion for choosing contact patterns or recovery policies that keep R0 below one.
  • The same rescaling technique may yield analogous threshold conditions for other network compartmental models such as SEIR.

Load-bearing premise

The network is fixed and connected and the rates of infection, recovery, and loss of immunity remain constant and proportional to the number of contacts.

What would settle it

For any chosen connected network and set of rates, compute R0; if R0 exceeds 1, integrate the ODE system from a small positive initial infection vector and verify that the infected fractions converge to a strictly positive steady-state vector independent of the precise starting values.

Figures

Figures reproduced from arXiv: 2604.21065 by Giacomo Como, Giulia Gatti.

Figure 1
Figure 1. Figure 1: Interaction network in Section V. Each link’s thick [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Time evolution of the infected (red) and recovered [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
read the original abstract

We study the Suscectible-Infected-Recovered-Susceptible (SIRS) epidemic model on deterministic networks. For connected but otherwise general interaction patterns and heterogeneous recovery and loss-of-immunity rates, we identify a fundamental parameter R_0 (the basic reproduction number), which fully characterizes the qualitative dynamic behavior of the system. This parameter is the dominant eigenvalue of a rescaled version of the interaction matrix, whose rows are normalized by the corresponding recovery rates. We prove that a transcritical bifurcation occurs as R_0 crosses the threshold value 1. Specifically, we show that, if R_0 does not exceed 1, then the disease-free equilibrium is globally asymptotically stable, whereas, if R_0 is larger than 1, then the disease-free equilibrium is unstable and there exists a unique endemic equilibrium, which is asymptotically stable. As a byproduct of our analysis, we also identify key monotonicity properties of the dependence of the endemic equilibrium on the model parameters (the interaction matrix as well as the recovery rates and the loss-of-immunity rates) and obtain a distributed iterative algorithm for its computation, with provable convergence guarantees. Our results extend existing ones available in the literature for network SIRS epidemic models with rank-one interaction matrices and homogeneous recovery rates (including the single homogeneous population SIRS epidemic model).

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 3 minor

Summary. The paper examines the SIRS epidemic model on deterministic networks with general connected interaction patterns and heterogeneous recovery and loss-of-immunity rates. It defines a basic reproduction number R_0 as the dominant eigenvalue of the row-rescaled interaction matrix (normalized by recovery rates). The main results are proofs that the disease-free equilibrium is globally asymptotically stable if R_0 ≤ 1, and that if R_0 > 1, the disease-free equilibrium is unstable and there exists a unique asymptotically stable endemic equilibrium, established through transcritical bifurcation analysis. The paper also derives monotonicity properties of the endemic equilibrium with respect to model parameters and presents a distributed iterative algorithm for computing it with convergence guarantees. These results generalize previous findings for rank-one matrices and homogeneous rates.

Significance. If the proofs are correct, this manuscript makes a significant contribution by providing a complete threshold characterization for SIRS dynamics on arbitrary connected networks with heterogeneity in rates. The extension beyond rank-one interaction matrices using matrix theory and monotonicity is noteworthy. The identification of R_0 as the key parameter and the provision of a distributed algorithm enhance both theoretical understanding and practical computation in networked epidemic models. This could influence future work on more general compartmental models on graphs.

major comments (2)
  1. [§3.1] §3.1 (global stability proof): The Lyapunov function for GAS of the disease-free equilibrium when R_0 ≤ 1 is constructed after row-rescaling by recovery rates; the derivative must be shown to remain non-positive when heterogeneous loss-of-immunity rates are present, as these rates appear in the SIRS equations but are not part of the rescaling.
  2. [§4.2] §4.2 (uniqueness of endemic equilibrium): The fixed-point argument establishing uniqueness for R_0 > 1 relies on monotonicity and irreducibility; it is not immediately clear whether the contraction or ordering properties continue to hold if loss-of-immunity rates approach zero, which would reduce the system toward an SIR limit.
minor comments (3)
  1. [Abstract] Abstract: 'Suscectible' is a typographical error and should read 'Susceptible'.
  2. [§5] The distributed algorithm in §5 is stated to converge, but no explicit convergence rate or iteration bound is given; adding this would improve practical utility.
  3. [§2] Notation for the rescaled matrix (e.g., the precise definition of the row-normalized operator) could be introduced earlier and used consistently to aid readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the positive and constructive report. The comments identify opportunities to strengthen the clarity of the proofs in Sections 3.1 and 4.2. We address each point below and will incorporate the requested clarifications in the revised manuscript.

read point-by-point responses

  1. Referee: [§3.1] §3.1 (global stability proof): The Lyapunov function for GAS of the disease-free equilibrium when R_0 ≤ 1 is constructed after row-rescaling by recovery rates; the derivative must be shown to remain non-positive when heterogeneous loss-of-immunity rates are present, as these rates appear in the SIRS equations but are not part of the rescaling.

    Authors: We appreciate the referee drawing attention to this detail. The Lyapunov function is constructed on the rescaled infected states and its derivative is evaluated along the full SIRS dynamics. The loss-of-immunity rates enter only the susceptible and recovered equations; they do not appear in the infected-state equations. Consequently, when the derivative is computed at or near the disease-free equilibrium (where the susceptible fraction equals one), the loss-of-immunity terms cancel or remain non-positive and do not affect the sign of the quadratic form that yields non-positivity for R_0 ≤ 1. We will revise §3.1 to display the complete derivative expression explicitly, confirming that heterogeneity in the loss-of-immunity rates preserves the non-positivity result. revision: yes

  2. Referee: [§4.2] §4.2 (uniqueness of endemic equilibrium): The fixed-point argument establishing uniqueness for R_0 > 1 relies on monotonicity and irreducibility; it is not immediately clear whether the contraction or ordering properties continue to hold if loss-of-immunity rates approach zero, which would reduce the system toward an SIR limit.

    Authors: We thank the referee for this observation on the limiting regime. The fixed-point map whose monotonicity and ordering properties are used for uniqueness depends only on the interaction matrix and the recovery rates; the loss-of-immunity rates appear as a strictly positive additive perturbation that preserves monotonicity and the contraction property for any positive value. As the loss-of-immunity rates tend to zero, the endemic equilibrium of the SIRS system converges to the unique endemic equilibrium of the corresponding SIR system (whose uniqueness is already established in the literature under the same R_0 > 1 condition). We will add a short remark in §4.2 that records this uniform validity of the ordering argument and the convergence to the SIR limit. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper explicitly defines R_0 as the dominant eigenvalue of the row-rescaled interaction matrix (with rows normalized by heterogeneous recovery rates) directly from the model parameters and network structure. It then derives the threshold behavior, transcritical bifurcation at R_0=1, global asymptotic stability of the disease-free equilibrium for R_0 ≤ 1, and existence/uniqueness/stability of the endemic equilibrium for R_0 > 1 using standard dynamical systems arguments (monotonicity, Perron-Frobenius on irreducible matrices from connectedness, and bifurcation theory). These proofs are independent of the definition and do not reduce the claimed results to fitted inputs, self-definitional loops, or load-bearing self-citations; the extension of prior rank-one cases is noted but not used as the sole justification for the general case. The derivation is self-contained against the stated assumptions.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The analysis rests on standard results from dynamical systems (transcritical bifurcation, Lyapunov stability) and graph theory (connectedness, spectral properties of nonnegative matrices). No new entities are postulated. The rescaled matrix is constructed directly from the given parameters without additional fitted constants.

axioms (2)
  • domain assumption The network is connected and the interaction matrix is nonnegative.
    Invoked to guarantee the dominant eigenvalue is well-defined and positive and to ensure the bifurcation behavior applies globally.
  • standard math Standard results on transcritical bifurcations and global asymptotic stability for compartmental epidemic models hold under the stated rate assumptions.
    Used to conclude stability from the sign of R_0 - 1.

pith-pipeline@v0.9.0 · 5540 in / 1435 out tokens · 26483 ms · 2026-05-09T23:09:39.830247+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

25 extracted references · 25 canonical work pages

  1. [1]

    A contribution to the mathematical theory of epidemics,

    W. O. Kermack and A. G. McKendrick, “A contribution to the mathematical theory of epidemics,”Proceedings of the Royal Society of London. Series A, vol. 115, no. 772, pp. 700–721, 1927

    work page 1927

  2. [2]

    Contributions to the mathe- matical theory of epidemics. ii. the problem of endemicity,

    W. O. Kermack and A. G. McKendrick, “Contributions to the mathe- matical theory of epidemics. ii. the problem of endemicity,”Proceed- ings of the Royal Society A, vol. 138, pp. 55–83, 1932

    work page 1932

  3. [3]

    R. M. Anderson and R. M. May,Infectious Diseases of Humans: Dynamics and Control. Oxford: Oxford University Press, 1991

    work page 1991

  4. [4]

    Diekmann and J

    O. Diekmann and J. A. P. Heesterbeek,Mathematical Epidemiology of Infectious Diseases: Model Building, Analysis and Interpretation. Wiley, 2000

    work page 2000

  5. [5]

    Some equations modelling growth pro- cesses and gonorrhea epidemics,

    K. L. Cooke and J. A. Yorke, “Some equations modelling growth pro- cesses and gonorrhea epidemics,”Mathematical Biosciences, vol. 16, no. 1-2, pp. 75–101, 1973

    work page 1973

  6. [6]

    Asymptotic behavior in a deterministic epidemic model,

    H. W. Hethcote, “Asymptotic behavior in a deterministic epidemic model,”Bulletin of Mathematical Biology, vol. 35, pp. 607–614, 1973

    work page 1973

  7. [7]

    Sexually transmitted diseases and sexual behavior: Insights from mathematical models,

    G. P. Garnett and R. M. Anderson, “Sexually transmitted diseases and sexual behavior: Insights from mathematical models,”The Journal of Infectious Diseases, vol. 174, pp. S150–S161, 10 1996

    work page 1996

  8. [8]

    Influence of nonlinear incidence rates upon the behavior of SIRS epidemiological models,

    W.-m. Liu, S. A. Levin, and Y . Iwasa, “Influence of nonlinear incidence rates upon the behavior of SIRS epidemiological models,” Journal of Mathematical Biology, vol. 23, no. 2, pp. 187–204, 1986

    work page 1986

  9. [9]

    An SIRS model with a nonlinear incidence rate,

    Y . Jin, W. Wang, and S. Xiao, “An SIRS model with a nonlinear incidence rate,”Chaos, Solitons & Fractals, vol. 34, pp. 1482–1497, 12 2007

    work page 2007

  10. [10]

    Lyapunov functions and global stability for SIR, SIRS, and SIS epidemiological models,

    A. Korobeinikov and G. Wake, “Lyapunov functions and global stability for SIR, SIRS, and SIS epidemiological models,”Applied Mathematics Letters, vol. 15, no. 8, pp. 955–960, 2002

    work page 2002

  11. [11]

    A deterministic model for gonorrhea in a nonhomogeneous population,

    A. Lajmanovich and J. A. Yorke, “A deterministic model for gonorrhea in a nonhomogeneous population,”Mathematical Biosciences, vol. 28, no. 3-4, pp. 221–236, 1976

    work page 1976

  12. [12]

    Monotone dynamical systems,

    M. Hirsch and H. Smith, “Monotone dynamical systems,” vol. 2 of Handbook of Differential Equations: Ordinary Differential Equations, Chapter 4, pp. 239 – 357, North-Holland, 2006

    work page 2006

  13. [13]

    Epidemic processes in complex networks,

    R. Pastor-Satorras, C. Castellano, P. Van Mieghem, and A. Vespig- nani, “Epidemic processes in complex networks,”Reviews of Modern Physics, vol. 87, pp. 925–979, 2015

    work page 2015

  14. [14]

    Modeling, estimation, and analysis of epidemics over networks: An overview,

    P. E. Par ´e, C. L. Beck, and T. Bas ¸ar, “Modeling, estimation, and analysis of epidemics over networks: An overview,”Annual Reviews in Control, vol. 50, pp. 345–360, 2020

    work page 2020

  15. [15]

    On the dynamic behavior of the network SIR epidemic model,

    M. Alutto, L. Cianfanelli, G. Como, and F. Fagnani, “On the dynamic behavior of the network SIR epidemic model,”IEEE Transactions on Control of Network Systems, vol. 12, no. 1, pp. 177–189, 2025

    work page 2025

  16. [16]

    Analysis and control of epidemics: A survey of spreading processes on complex networks,

    C. Nowzari, V . M. Preciado, and G. J. Pappas, “Analysis and control of epidemics: A survey of spreading processes on complex networks,” IEEE Control Systems Magazine, vol. 36, no. 1, pp. 26–46, 2016

    work page 2016

  17. [17]

    On the dynamics of deterministic epidemic propagation over networks,

    W. Mei, S. Mohagheghi, S. Zampieri, and F. Bullo, “On the dynamics of deterministic epidemic propagation over networks,”Annual Reviews in Control, vol. 44, pp. 116–128, 2017

    work page 2017

  18. [18]

    Mathematical epidemiology: Past, present, and future,

    F. Brauer, “Mathematical epidemiology: Past, present, and future,” Infectious Disease Modelling, vol. 2, no. 2, pp. 113–127, 2017

    work page 2017

  19. [19]

    Analysis, prediction, and control of epidemics: A survey from scalar to dynamic network models,

    L. Zino and M. Cao, “Analysis, prediction, and control of epidemics: A survey from scalar to dynamic network models,”IEEE Circuits and Systems Magazine, vol. 21, no. 2, pp. 4–24, 2021

    work page 2021

  20. [20]

    Analysis of epidemic spreading of an SIRS model in complex heterogeneous networks,

    C. Li, C. Tsai, and S. Yang, “Analysis of epidemic spreading of an SIRS model in complex heterogeneous networks,”Communications in Nonlinear Science and Numerical Simulations, vol. 19, pp. 1042– 1054, 2014

    work page 2014

  21. [21]

    Global stability and optimal control of an SIRS epidemic model on heterogeneous networks,

    L. Chen and J. Sun, “Global stability and optimal control of an SIRS epidemic model on heterogeneous networks,”Physica A: Statistical Mechanics and its Applications, vol. 410, pp. 196–204, 2014

    work page 2014

  22. [22]

    Layered SIRS model of information spread in complex networks,

    Y . Zhang and D. Pan, “Layered SIRS model of information spread in complex networks,”Applied Mathematics and Computation, vol. 411, p. 126524, 2021

    work page 2021

  23. [23]

    Epidemic threshold for the SIRS model on the networks,

    M. A. Saif, “Epidemic threshold for the SIRS model on the networks,” Physica A: Statistical Mechanics and its Applications, vol. 535, p. 122251, 2019

    work page 2019

  24. [24]

    Global stability of a network-based SIRS epidemic model with nonmonotone incidence rate,

    L. Liu, X. Wei, and N. Zhang, “Global stability of a network-based SIRS epidemic model with nonmonotone incidence rate,”Physica A, vol. 515, pp. 587–599, 2019

    work page 2019

  25. [25]

    Berman and R

    A. Berman and R. J. Plemmons,Nonnegative Matrices in the Math- ematical Sciences. Classics in Applied Mathematics, Philadelphia: SIAM, 1994

    work page 1994