An Age-Dependent Model for Dengue Transmission: Analysis and Comparison to Field Data from Semarang, Indonesia

K.P. Wijaya; N.C. Ganegoda; T. G\"otz

REVIEW 4 major objections 5 minor 18 references

An Age-Dependent Model for Dengue Transmission: Analysis and Comparison to Field Data from Semarang, Indonesia

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Age-structured dengue model gets a unique endemic equilibrium and a convergent iteration.

desk verdict The math is sound and the fixed-point convergence result is real, but the Semarang data fit compares a steady-state prevalence curve to annual incidence counts, which is a genuine flaw in the empirical validation. read the letter →

arxiv 1908.09256 v1 pith:GAT5HQJK submitted 2019-08-25 q-bio.PE math.DS

N.C. Ganegoda , T. Götz , K.P. Wijaya This is my paper · ORCID

classification q-bio.PEmath.DS MSC 92D3045G10

keywords age-structuredepidemicmodeldenguetransmissionbasicreproductivenumberfixed-pointiterationintegro-differentialequationparameterestimationSemaranghospitalizationagedistribution

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper establishes that the equilibrium age profile of an age-structured SI dengue model is governed by a scalar fixed-point equation, and that this equation has a unique nontrivial solution exactly when the basic reproductive number $R_0$ exceeds one. It also proves that a simple iteration converges to that solution for $1

What carries the argument

The central object is the scalar fixed-point equation for $Q$, the normalized infected proportion of the host population. It reduces the PDE equilibrium problem to a one-dimensional root-finding problem, $F(Q)=0$, where $F(Q)=1-\int\int_{\Omega} k(x,s)e^{-Q\Lambda(x,s)}\,d(s,x)$. The kernel $k(x,s)=p(x)\theta(s)e^{-\gamma(x-s)}/N$ and the accumulated exposure $\Lambda(x,s)=\int_s^x \theta(t)\,dt$ are what make the age structure visible in $R_0$; the monotonicity and concavity of $F$ supply existence and uniqueness, while the iteration map $T(Q)=Q(1-F(Q))$ supplies convergence. This machinery carries the entire argument from threshold characterization to the numerical parameter fits.

What would settle it

Use the optimized transmission curve from the six-year fit to simulate the model under each year's observed incidence forcing without averaging; if the resulting age distribution diverges from the reported individual-year data while the averaged data match, the steady-state-as-incidence identification fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that the endemic equilibrium of the reduced host-only model is determined by the scalar $Q$, the normalized infected proportion, through the fixed-point equation $Q = Q \int\int_{\Omega} k(x,s)e^{-Q\Lambda(x,s)} \, d(s,x)$, where $k(x,s) = p(x)\theta(s)e^{-\gamma(x-s)}/N$ encodes population structure, age-dependent transmission, and loss of immunity. Defining $R_0 = \int\int_{\Omega} k(x,s)\,d(s,x)$, the authors prove that for $R_0\leq 1$ only the disease-free equilibrium exists, while for $R_0>1$ a unique $Q^*\in(0,1)$ exists; moreover, the iteration $Q^{(j)}=T(Q^{(j-1)})$ is locally convergent to $Q^*$ when $1<R_0<2e$, with $Q=0$ locally repelling. With the fitted Gaussian-plus-background transmission rate, the stationary infected fraction $i(x)=w(x)p(x)$ reproduces the averaged Semarang age distribution, and the SIR variant slightly improves the fit.

Load-bearing premise

The load-bearing premise is that the model's stationary infected fraction is proportional, without an explicit scale factor, to the observed yearly hospitalized-case counts; if that proportionality fails, the fitted transmission parameters do not estimate the intended age-dependent infection rate.

Editorial extensions

If this is right

If $R_0 \leq 1$, the model permits only the disease-free equilibrium, so endemic dengue persistence in this age-structured setting requires the reproductive number to exceed one.
For $1<R_0<2e$, the fixed-point iteration is locally convergent, so the endemic age distribution can be computed reliably whenever fitted parameters fall in that range.
The optimized transmission rate is maximal at birth and decreases with age, which the model identifies as the reason hospitalized-case counts are concentrated in younger age classes.
The SIR extension lowers the $L^2$ error from 135.55 to 128.48, showing that adding a recovered class improves agreement with the averaged hospitalization data.
For a constant transmission rate and linearly declining population, the asymptotic solution $z\simeq k-2$ recovers the classic SI equilibrium value $1-\gamma/\theta_0$ in the large-$A\theta_0$ limit, linking the age-structured model to the familiar age-independent result.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The data are annual hospitalization counts while the model is a stationary prevalence profile; fitting without an explicit age-dependent reporting or hospitalization fraction implicitly assumes those two are proportional, an assumption the paper does not test.
The theorem's convergence range is capped at $R_0<2e$ by a bounding argument rather than by a sharp condition; because the paper does not report fitted $R_0$ values, the margin between the fitted regime and this cap is unknown.
A direct extension would be to fit the age-dependent reporting fraction as an unknown parameter; the resulting transmission curve would likely shift if hospitalization probability varies strongly with age.
The same existence-and-convergence framework should extend to age-dependent recovery rates or to an SIR structure with age-dependent mortality, since the kernel and monotonicity arguments only require continuity and positivity of the coefficient functions.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague—

The paper has a solid analytical core and a shaky empirical wrapper. If you read it for the math, you'll find a clean, self-contained existence and convergence analysis for the equilibrium of an age-structured SI model. The genuinely new bits are the local convergence theorem for the fixed-point iteration (with the R0 < 2e bound) and the asymptotic expansion in the simplified case. The existence theorem is classical in spirit, but the presentation is clear and the proofs are short. Credit where it's due: the authors know the age-structured literature (Inaba, Iannelli, et al.) and don't oversell the novelty. The sign error in Eq. (21) and the label confusion (alpha vs sigma in Table 1) are fixable.

The soft spot is the data comparison, and it's not minor. The model's i(x) is the steady-state number of infected (prevalence). The Semarang data are annual hospitalizations (incidence). The paper fits to i(x) without any justification that the two are proportional across ages. The stress-test note is right: in an age-structured model, incidence and prevalence are not proportional when w'(x) or mortality vary with age. A constant hospitalization rate would scale the whole curve but wouldn't change the shape, so if the comparison is only about shape, the mismatch might not be fatal. But the paper doesn't say that, and the absolute L2 errors (135.55 etc.) suggest they're comparing raw counts. The fitted peak at newborns (xp < 0) is suspicious for this reason. This is the load-bearing empirical support for the model's 'validity,' and it's an in-sample fit with no uncertainty quantification or out-of-sample check. That part needs rethinking.

The theoretical section holds up. The iteration convergence proof is a modest but real contribution, and the asymptotic analysis is a nice sanity check. The citation pattern is honest. This is not a breakthrough, but it's a competently done contribution to mathematical epidemiology that deserves a serious referee.

My recommendation: send it to peer review, but make sure the referees push on the prevalence/incidence identification. The authors should either show that the normalized age profiles are robust to the distinction, or add a hospitalization/reporting rate and do a proper parameter-uncertainty analysis. Without that fix, the data fit shouldn't be reported as validation.

— [Your name]

Referee Report

4 major / 5 minor

Summary. This paper proposes an age-structured, vector-borne dengue model and reduces the vector dynamics by a time-scale argument to an age-structured SI-type host model. The endemic equilibrium is characterized by a scalar fixed-point problem for the total infected fraction Q, with infected age profile w(x) solving an ODE. The main analytical results are: if R0 <= 1 only the disease-free equilibrium exists, while if R0 > 1 a unique endemic equilibrium exists (Theorem 6); a fixed-point iteration converges locally for 1 < R0 < 2e (Theorem 10). The paper also derives an approximate closed-form solution for constant transmission and fits the model to hospitalized dengue case data from Semarang by least squares, comparing SI and SIR variants.

Significance. The analytical core is a solid, self-contained contribution: Theorem 6 follows from the monotonicity of F in Lemma 5, and Theorem 10 gives a concrete contraction range in terms of R0. The authors also correctly note that their fitted parameters place R0 within the proved convergence range. However, the empirical validation, which is advertised as a central contribution, is not convincing. The model output w(x)p(x) is an equilibrium prevalence, whereas the Semarang data are annual hospitalized-case counts, i.e., incidence; no hospitalization or reporting rate is introduced. In addition, the parameters are estimated on the same data used for validation. These issues directly affect the fitted parameters and the headline L2 errors, so the applied claims need substantial revision even though the theoretical results appear sound.

major comments (4)

[Sec. 4.1-4.2, Eq. (5), Figs. 6-7] The quantity compared to the data is w(x)p(x)=i(x), the stationary prevalence of infected individuals, but the data are annual numbers of hospitalized cases, which are incidence flows. In the equilibrium of the age-structured SI model, the age-specific annual incidence of new infections is theta(x)(1-w(x))Q p(x)/N with Q defined in Eq. (5b), not i(x). These two curves are not proportional because w satisfies the differential equation in (5a), so the age profiles differ by the w'(x) term. The least-squares problem in Sec. 4.2 minimizes ||w-data||, so a prevalence stock is being fitted to incidence data. This can bias the estimated parameters, including the reported negative xp that places the transmission peak at newborns. The authors should either derive the model's incidence curve and fit that curve, include an age-dependent hospitalization or reporting rate, or explicitly justify and verify a proportionality between prevalence and annual hospitalized cases.
[Sec. 4.2, Table 1] The parameters are chosen by minimizing the L2 distance to the same averaged data set that is then shown as agreement in Figs. 6 and 7. This is in-sample fitting rather than validation. The instability of the estimates is visible in Table 1: across the six yearly fits, xp ranges from -11.42 to 5.07 and theta1 from 1.23 to 2.70, yet only the average is used to support the model. A cross-validation or independent-year prediction, together with an uncertainty or identifiability analysis, is needed before the model can be described as validated.
[Sec. 4.2-4.3 vs. Eq. (5b), Eq. (11)] There is a normalization inconsistency in the definition of Q. The fixed point is defined as Q = (1/N) integral p(y)w(y) dy in Eq. (5b) and Eq. (11), but the constraint in Sec. 4.2 is written with theta(1-w)Q/N, and Sec. 4.3 defines Q = integral_0^A p(y)w(y) dy without the factor 1/N. Since N is approximately 1.6 x 10^6, this is not a harmless rescaling: if the numerical implementation follows the written equations, the force of infection differs by a factor of N. The authors should state exactly which normalization was implemented and verify that the reported theta values are consistent with the model equations as written.
[Sec. 4.3] The SIR model is fit to the same incidence data using the same prevalence-incidence identification, so the improved L2 error of 128.48 does not provide independent support for the method. Furthermore, the fixed-point iteration and convergence theorem were proved only for the SI model; the SIR equilibrium equations are not analyzed in Sec. 3, so the SIR comparison lacks the theoretical backing claimed for the overall framework. The SIR section should either be supported by an analogous analytical result or explicitly presented as a purely numerical comparison.

minor comments (5)

[Eq. (8) and Table 1] The width parameter of the Gaussian transmission rate is denoted sigma in Eq. (8), but Table 1 and the surrounding text use alpha (e.g., the column header 'alpha' and the discussion after Table 1). Please use one symbol consistently.
[Eq. (1), Eq. (5)] The model in Eqs. (1) and (5) is SIS-like, because infected individuals return to the susceptible class at rate gamma; calling it an SI model is confusing, especially when the SIR model in Sec. 4.3 is presented as an extension.
[Eq. (21)] The prefactor [1 - A gamma/2 - A theta0] in Eq. (21) appears inconsistent with the value alpha/(gamma+alpha) obtained from Eq. (20). With the fitted parameter values (A=100, theta0=0.1821, gamma=0.6412) the prefactor is negative, which cannot be a valid endemic prevalence coefficient. Please check the algebra leading to this displayed formula.
[Fig. 5] The y-axis label in Fig. 5 says 'Percentage of infected cases', but the data are percentages of hospitalized cases. Please re-label to avoid conflating infection with hospitalization.
[Sec. 3.3, Eq. (15)] The formula for f(delta) in Eq. (15) is stated without derivation; adding one line of integration would make the section easier to check.

Circularity Check

1 steps flagged · score 6.0 of 10

Empirical validation is an in-sample least-squares fit presented as a prediction; the analytical Theorems 6 and 10 are independent.

fitted input called prediction [Section 4.2 (Parameter estimation), Eq. (5), Table 1, Fig. 6; cf. Section 4.3, Table 2, Fig. 7]
"In a consecutive step, we will use the age-structured SI model (5) to predict the age distribution of dengue cases. For that purpose we need to identify the parameters c := (θ0, θ1, xp, σ) shaping the transmission rate θ, as well as the rate of loss of immunity γ > 0. We estimate these model parameters, such that the model solution w and the observed data agree in the least-squared sense: min_{(c,γ)∈R4×R+} ‖w − data‖2."

The 'predicted' age distribution is the direct object of the least-squares objective: the parameters c and γ are chosen to minimize ‖w − data‖². The subsequent agreement between w and the Semarang data (L2 error 135.55 for SI, 128.48 for SIR) is the minimized in-sample residual, not an independent prediction. The same six-year averaged data are used both to identify all five parameters and to 'validate' the model; no held-out data or out-of-sample test is performed. Thus the empirical validation claim reduces by construction to the fitting procedure. The mathematical theorems in Section 3 remain self-contained, but the applied validation is a fitted input presented as a prediction.

full rationale

The mathematical core (Sections 3.1–3.2) is self-contained: Lemmas 4, 5, 8 and 9 provide the estimates that yield Theorem 6 (existence and uniqueness of Q*) and Theorem 10 (local convergence of the fixed-point iteration), using only the stated assumptions on θ, p, and γ. There is no self-citation chain or imported uniqueness theorem; references [5]–[13] are background, not load-bearing, and no claim rests on the authors' prior work. The circularity is confined to the empirical section. Section 4.2 explicitly says it will 'predict' the age distribution and then identifies c and γ by minimizing the squared distance between w and the Semarang data. Figures 6–7 and Tables 1–2 then report the minimized residual as agreement and validation. That is an in-sample fit: the observed data are the fitting target, so the reported L2 errors are forced by construction rather than independent predictions. This matches the fitted-input-called-prediction pattern. A separate limitation, not itself circularity, is the identification of the equilibrium infected stock w(x)p(x) with annual hospitalized-case incidence without a hospitalization or reporting rate; this can bias the fitted parameters but is an empirical misspecification concern, not a circular derivation. Overall score 6 because the central analytical derivations are independent, but the empirical validation claim partially reduces to the fitting procedure.

Assumptions & free parameters 6 free parameters · 7 assumptions · 0 invented entities

No new biological entities are introduced. The effective infection rate theta(x) is a composite of vector and host parameters, not an entity. The main burden is a set of domain assumptions (steady state, vector elimination, prevalence-incidence identification) and five transmission and immunity parameters fitted to the same data used for validation.

free parameters (6)

theta0 = 0.1821 (average 2009-2014; per-year 0.0919 to 0.2226)
Background transmission rate in Eq. (8), estimated by least squares on Semarang data in Section 4.2.
theta1 = 2.1135 (average 2009-2014)
Gaussian peak height in Eq. (8), estimated by least squares.
xp = -0.5209 (average; range -11.42 to 5.07 across years)
Location of peak transmission in Eq. (8), estimated by least squares; highly variable across single-year fits.
sigma (labeled alpha in Table 1) = 10.3164 (average 2009-2014)
Gaussian width in Eq. (8), estimated by least squares; notation is inconsistent in the paper.
gamma = 0.6412 (average 2009-2014)
Rate of loss of immunity in the SI/SIR model, estimated by least squares.
k0, k1, k2 = 0.0181*N, 0.0984, 55.3
Logistic population age-distribution parameters fitted to Indonesia population pyramid data in Section 2.3, Eq. (6).

assumptions (7)

domain assumption Host population is at a stationary age distribution p(x) with constant total population N.
Section 2.1, after Eq. (2): the paper considers the host population in steady state and assumes constant total population.
domain assumption Vector population is eliminated by quasi-steady-state with J/rho much less than 1, giving v* approximately (lambda/rho^2)J.
Section 2.2, Eq. (4) and surrounding text; load-bearing for the host-only SI reduction.
domain assumption No vertical transmission; all newborns are susceptible.
Section 2.1: the epidemic assumes no vertical transmission.
ad hoc to paper Transmission rate has the parametric form theta(x) = theta0 + theta1 exp(-(x-xp)^2/(2 sigma^2)).
Eq. (8), introduced as a model inspired by the data curve; not derived from mechanism.
domain assumption Infected hosts lose immunity at a uniform rate gamma independent of age.
Section 2.1: infected hosts of any age lose immunity at a uniform rate gamma.
domain assumption Hospitalized hosts are isolated, so vector transmission from them is negligible (sigma(y) approximately 0).
Section 2.2, used to justify the Taylor expansion of v*.
domain assumption Stationary prevalence can be compared directly to annual hospitalized incidence.
Section 4.1-4.2 and Figures 5-6; this premise is unflagged and load-bearing for the empirical validation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of An Age-Dependent Model for Dengue Transmission: Analysis and Comparison to Field Data from Semarang, Indonesia." pith.science (2026). https://pith.science/paper/GAT5HQJK

@misc{pith2026190809256,
  author       = {Pith},
  title        = {Pith review of: An Age-Dependent Model for Dengue Transmission: Analysis and Comparison to Field Data from Semarang, Indonesia},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GAT5HQJK}},
  note         = {Machine review of arXiv:1908.09256}
}

read the original abstract

Medical statistics reveal a significant dependence of hospitalized dengue patient on the patient's age. To incorporate an age-dependence into a mathematical model, we extend the classical ODE system of disease dynamics to a PDE system. The equilibrium distribution is then determined by the fixed points of resulting integro-differential equations. In this paper we use an extension of the concept of the basic reproductive number to characterize parameter regimes, where either only the disease-free or an endemic equilibrium exists. Using rather general and minimal assumptions on the population distribution and on the age-dependent transmission rate, we prove the existence of those equilibria. Furthermore, we are able to prove the convergence of an iteration scheme to compute the endemic equilibrium. To validate our model, we use existing data from the city of Semarang, Indonesia for comparison and to identify the model parameters.

Figures

Figures reproduced from arXiv: 1908.09256 by the authors.

**Figure 1.** Data (year 2016; summed over both sexes) for the age pyramid (blue) and fitted parametric model (6) (red). 2.3. Models for the total population and infection rate. The data of hospitalized cases are recorded from the city of Semarang, Indonesia. To accompany the data assimilation using these data, we first fit a parametric model (6) p(x; k0, k1, k2) = k0 1 + e k1(x−k2) to available population data with age structur… view at source ↗

**Figure 2.** Typical shape of the stationary age distribution of the infected class (solid blue) and the underlying transmission rate θ (dashed red). 3. Fixed point equation for the equilibrium In this section we analyze the solvability of the equilibrium problem (5). To prove the existence of equilibria and convergence for an according iteration scheme in a rather generalized setting, we meet the following minimal assumptions o… view at source ↗

**Figure 3.** Contour plot for R0 depending on γ and xp. For R0 > 1, the fixed point problem (5) admits a non–trivial solution. 3.3. Approximate analytical solution. Here, we consider an approximate solution for a simplified version of the fixed point problem (5). We assume a transmission rate θ(x) ≡ θ0 independent of the age x and a linear behaviour of the population p(x) = N0 (1 − x/A). In this case, N = R A 0 p(x) dx = N0A 2… view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Solutions z vs. k of the full equation (16) (solid blue), the cubic equation (17) written for z instead of y (dashed red) and the asymptotic expansion (19) (dashed green). For k ≫ 1, the asymptotic expansion z ≃ k − 2 − 2/k provides a good approximation of the non-tri…

**Figure 5.** Figure 5: Hospitalized dengue cases in Semarang from 2009 to 2014. Shown is the percentage of cases (relative to the total number of infections) within each of the 14 age classes. 4.2. Parameter estimation. In a consecutive step, we will use the age-structured SI model (5) to pr…

**Figure 6.** Figure 6: Comparison of data (average of observed cases in the years 2009–2014) and simulation for stationary state of infected class. 4.3. SIR model. In a last step, we extend the previous SI model (5) to an SIR model w ′ = −γw + θ(x; c) · (1 − w − z) · Q N , w(0) = 0, z ′ = γw…

**Figure 7.** Figure 7: Comparison average data and simulations using the SI and SIR model. Parameter θ0 θ1 xp α γ ku − datak Optimal SI 0.1821 2.1135 −0.5209 10.3164 0.6412 135.55 Optimal SIR 0.1807 2.6877 −5.8795 12.7946 0.6511 128.4848 [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

18 extracted references · 18 canonical work pages

[1]

Coale, Growth and Structure of Human Populations: A Ma thematical Investigation, Princeton University Press, 1972

A. Coale, Growth and Structure of Human Populations: A Ma thematical Investigation, Princeton University Press, 1972

work page 1972
[2]

Hoppenstaedt, Mathematical Theories of Populations , Society for Industrial and Applied Mathematics, 1975

F. Hoppenstaedt, Mathematical Theories of Populations , Society for Industrial and Applied Mathematics, 1975

work page 1975
[3]

Dietz, D

K. Dietz, D. Schenzle, Proportionate mixing models for a ge-dependent infection transmission, Journal of Mathematical Biology 22 (1) (1985) 117–120

work page 1985
[4]

J. Metz, O. Diekmann, The Dynamics of Physiologically St ructured Populations, Springer Verlag, 1986

work page 1986
[5]

de Le´ on, L

C. de Le´ on, L. Esteva, A. Korobeinikov, Age-dependency in host-vector models - The global analysis., Applied Mathematics and Computation 243 (2014) 969–981

work page 2014
[6]

K. Rock, D. W ood, M. Keeling, Age- and bite-structured mo dels for vector-borne diseases, Epidemics 12 (2015) 20–29

work page 2015
[7]

W ang, Y

X. W ang, Y. Chen, S. Liu, Dynamics of an age-structured ho st-vector model for malaria transmission, Mathematical Methods in the Applied Science s 41 (5) (2018) 1966–1987

work page 2018
[8]

Busenberg, K

S. Busenberg, K. Cooke, M. Iannelli, Endemic Thresholds and Stability in a Class of Age- Structured Epidemics, SIAM Journal on Applied Mathematics 48 (6) (1988) 1379–1395

work page 1988

Show all 18 references

[9]

Inaba, Threshold and stability results for an age-str uctured epidemic model, Journal of Mathematical Biology 28 (4) (1990) 1–25

H. Inaba, Threshold and stability results for an age-str uctured epidemic model, Journal of Mathematical Biology 28 (4) (1990) 1–25. 16 AGE–DEPENDENT DENGUE MODEL: ANALYSIS AND COMPARISON TO F IELD DATA

1990
[10]

Iannelli, F

M. Iannelli, F. Milner, A. Pugliese, Analytical and Num erical Results for the Age-Structured S-I-S Epidemic Model with Mixed Inter-Intracohort Transmi ssion, SIAM Journal on Mathe- matical Analysis 23 (3) (1992) 662–688

1992
[11]

Capasso, Mathematical structures of epidemic syste ms, Springer Verlag, 1993

V. Capasso, Mathematical structures of epidemic syste ms, Springer Verlag, 1993

1993
[12]

Inaba, H

H. Inaba, H. Sekine, A mathematical model for Chagas dis ease with infection-age-dependent infectivity, Mathematical Biosciences 190 (1) (2004) 39–6 9

2004
[13]

Franceschetti, A

A. Franceschetti, A. Pugliese, Threshold behaviour of a SIR epidemic model with age struc- ture and immigration., Journal of Mathematical Biology 57 ( 1) (2008) 1–27

2008
[14]

C. o. S. Health Oﬃce (Dinas Kesehatan), Dengue statisti cs, Private Communication (2014)

2014
[15]

Esteva, C

L. Esteva, C. Vargas, Analysis of a dengue disease trans mission model, Mathematical Bio- sciences 150 (2) (1998) 131–151

1998
[16]

A. Jain, U. Chaturvedi, Dengue in infants: an overview, FEMS Immunology & Medical Microbiology 59 (2) (2010) 119–130

2010
[17]

de Jong, M.C.M .and Diekmann, H

O. de Jong, M.C.M .and Diekmann, H. Heesterbeek, How doe s transmission of infection depend on population size?, in: D. Mollison (Ed.), Epidemic Models Their Structure and Relation to Data, The Newton Institute, Cambridge, 1995, pp . 84–94

1995
[18]

Pyramid.net

P. Pyramid.net. Population pyramid indonesia 2016 [on line] (2019) [cited 18.06.2019]

2019

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Coale, Growth and Structure of Human Populations: A Ma thematical Investigation, Princeton University Press, 1972

A. Coale, Growth and Structure of Human Populations: A Ma thematical Investigation, Princeton University Press, 1972

work page 1972

[2] [2]

Hoppenstaedt, Mathematical Theories of Populations , Society for Industrial and Applied Mathematics, 1975

F. Hoppenstaedt, Mathematical Theories of Populations , Society for Industrial and Applied Mathematics, 1975

work page 1975

[3] [3]

Dietz, D

K. Dietz, D. Schenzle, Proportionate mixing models for a ge-dependent infection transmission, Journal of Mathematical Biology 22 (1) (1985) 117–120

work page 1985

[4] [4]

J. Metz, O. Diekmann, The Dynamics of Physiologically St ructured Populations, Springer Verlag, 1986

work page 1986

[5] [5]

de Le´ on, L

C. de Le´ on, L. Esteva, A. Korobeinikov, Age-dependency in host-vector models - The global analysis., Applied Mathematics and Computation 243 (2014) 969–981

work page 2014

[6] [6]

K. Rock, D. W ood, M. Keeling, Age- and bite-structured mo dels for vector-borne diseases, Epidemics 12 (2015) 20–29

work page 2015

[7] [7]

W ang, Y

X. W ang, Y. Chen, S. Liu, Dynamics of an age-structured ho st-vector model for malaria transmission, Mathematical Methods in the Applied Science s 41 (5) (2018) 1966–1987

work page 2018

[8] [8]

Busenberg, K

S. Busenberg, K. Cooke, M. Iannelli, Endemic Thresholds and Stability in a Class of Age- Structured Epidemics, SIAM Journal on Applied Mathematics 48 (6) (1988) 1379–1395

work page 1988