Fractional modeling of hepatitis B virus transmission via heterosexual and homosexual contacts and its disability burden Authors: Kamel Guedri, Rahat Zarin, Basim M Makhdoum, Hatoon A Niyazi, Bandar M Fadhl, Olumuyiwa James Peter Published: 2025 ABSTRACT AbstractThe hepatitis B virus (HBV) is a major public health problem worldwide. Sexual intercourse, heterosexual and homosexual, is an important factor in HBV transmission in adults. This study develops a fractional-order mathematical model to describe the spread of HBV, using fractional calculus with a Mittag–Leffler kernel to account for memory effects and hereditary properties of biological systems. The adult population is divided into the groups of susceptible, infected, carrier, and recovered, on the basis of sex and sexual habits. The model incorporates HBV-related disability using gender-specific probabilities that capture long-term functional impairments among carriers. We derive the basic reproduction number (R0) and analyze the disease-free equilibrium (E0). Theoretical results include conditions ensuring existence and uniqueness of solutions and the Hyers–Ulam stability of the system. Sensitivity analysis of R0 identifies infection rates among same-sex partners, contact frequency, and immunity levels as major factors influencing HBV dynamics. Numerical simulations based on the generalized Mittag–Leffler kernel illustrate the model’s behavior over time. The findings suggest that strengthening immunization programs, reducing the number of carriers, and promoting safe sexual practices can effectively control HBV transmission. The study highlights the value of fractional-order models for capturing complex disease processes and evaluating long-term health outcomes within affected populations.Keywords: Memory effects, Hepatitis b virus, Disability burden, Sensitivity analysis, Nonlinear systems, Numerical simulation. Introduction Hepatitis B is a major world health problem caused by the hepatitis B virus (HBV), which primarily targets the liver. The disease has a wide variety of clinical pictures ranging from acute self-limiting disease to chronic complications of the liver, such as cirrhosis and hepatocellular carcinoma. Transmission of HBV occurs by contact with infected blood or by semen or other body fluids. High-risk behavior, e.g. multiple sexual partners and sexual intercourse in the absence of protection, particularly with those chronically infected with HBV, plays an important role in spreading HBV. HBV is extremely contagious as this virus is 50 to 100 times more infectious than HIV, and is much more stable environmentally than hepatitis C virus (see [1, 2]. Approximately 2 billion people worldwide have been infected by HBV with an annual mortality rate of almost 887,000 attributable to HBV complications in 2015 [3]. Despite advances made in examination, an effective cure for chronic HBV disease remains elusive. Consequently, vaccination with the hepatitis B vaccine forms the basis of all preventive measures and affording invaluable protection against HBV. Despite the development of anti-viral drugs capable of suppressing viral replication and reducing the inflammatory reaction in the liver, a definitive cure for chronic HBV remains elusive. Flow treatment approaches, such as nucleoside analogues and interferon therapy, aim to treat the infection and prevent possible disease development but these treatments are not curative. This reflects the prominence of counteractive measures as a viable anti-HBV strategy. Vaccination by means of hepatitis B immumization is a major cornerstone in the prevention of HBV. The vaccination application have been proved highly efficient in preventing HBV infection and its complications. General public vaccination programs incorporating hepatitis B antibodies have given rise to a conspicuous decrease in the prevalence of HBV, particularly in areas where the vaccination has been widely administered. This is especially the case in countries where the coverage with vaccination against HBV has been high. Here the incidence of HBV-related diseases, for instance, has been often greatly reduced in incidence, including childhood HCC [3]. Hepatitis B ranks in the main three irresistible illnesses with about 300,000 deaths associated to HBV-related ailment appendages for every year [4]. Data from the national hepatitis seroepidemiological survey in China (1992) suggested that there were about 120 million HBsAg carriers [5]. Further, hepatitis B is amongst the chief determinant of disease mortality in the country [6]. To reduce this public health burden, the China-GAVI Project (CGP) started vaccinating infants with hepatitis B vaccine in less developed areas such as western provinces and poor counties of central China in 2002. An estimated 3.8 million chronic HBV infections and 680,000 deaths were averted during the CGP (2003–2009) [7]. Retroactive routine vaccination of those under 15 years was then recommended in 2009. Although HBV infection has been substantially curtailed in children, the prevalence of this virus among adults is still very high. In those aged 20 years and older, the reported number of HBV case-patients increased by 22%, from 740,000 in 2004 to 903,000 in 2014 [8, 9]. The risk of chronic HBV carrier status is strongly influenced by the age at which infection occurs, with distinct transmission pathways in adults and children. Sexual transmission represents a significant route of HBV spread among adults. Heterosexual transmission has historically been identified as the primary mode of HBV dissemination in this demographic [10, 11]. Mathematical models have extensively explored heterosexual HBV transmission dynamics [9]. However, homosexual contact accounts for an estimated 10% of HBV cases, with male-to-male sexual transmission being particularly efficient [12]. The World Health Organization (WHO) indicates that MSM should undergo HBV testing as part of routine care because these individuals are at an increased risk of HEVs. Studies from the US and Europe as well as additional studies included from China has started to describe of transmission of HBV among MSM, but such studies are needed for the characterization in SW-China [13]. For example, the overall HBV prevalence among MSM was 7.7% in Kunming, a city located at southwestern China [14]. Women who have sex with women (WSW) There has been increasing awareness over the past decade of health issues and risks for WSW. HBV is more prevalent in males than females, but such evidence indicates that HBV transmission between women may be present. Risk factors among WSW include sexual relationships with men, multiple bisexual partners, and participation in contract marriages (unions between gay men and lesbian women) [15–17]. A sociological study in China revealed that 60–80% of WSW engaged in oral-genital contact, which presents a potential route for HBV transmission [17]. There is an increasing body of the medical literatures suggesting that chronic HBV-infection is an important cause for disability at population level. This effect is often measured using metrics such as YLDs and DALYs. The Global Burden of Disease Study 2019 estimated there were 316 million people who were chronically infected with HBV, resulting in around 555,000 deaths in this year alone; however, mortality data are not equivalent to the disease burden; chronic consequences including cirrhosis and hepatocellular carcinoma (HCC) will also lead to life-long morbidity and reduced quality of life [18, 19]. The DALY metric includes premature death and years spent living with functional limitation, thus reflecting the non-fatal burden of HBV infection including viral hepatitis sequelae in terms of chronic hepatic disorders, fatigue (loss of energy) and neurocognitive impairment [20]. In 2019, roughly 34.5% of age-standardized HBV DALYs were associated with modifiable risk factors including alcohol consumption, tobacco use, and high body mass index (BMI), demonstrating the influence of lifestyle behaviors on HBV-related disability [21]. This emphasizes the importance of preventive strategies in mitigating long-term disability among carriers. The global distribution of HBV-related disability is highly uneven; low- and middle-income regions experience greater DALY rates due to limited access to vaccination, testing, and treatment [18, 22]. Country-specific analyses have quantified these burdens more precisely, for instance, studies from Brazil and Iran report measurable YLD contributions from HBV-associated cirrhosis. In Brazil, chronic HBV and its sequelae attributed to approximately 57380DALYs (5.5 YLDs per 100,000) in 2008 of which the majority were due to premature mortality and a significant proportion arose from chronic disability [23]. Additionally, in 2013 about 1.1% of HBV-related DALYs could be attributed directly to injection drug use and shows how risk behaviors enhance disability in marginalized groups [24]. Together, these results confirm that HBV causes not only death but also long-term disability at the population level. From a modeling standpoint, YLD and DALY parameters translate carrier prevalence into estimated disability burden and facilitate needs-based planning for health and disability interventions. Numerous mathematical models have been formulated to investigate the transmission dynamics and control of HBV, representing substantial global research efforts [25–29]. For example, the influence of control measures and age structure on HBV spread in New Zealand was examined in [30], while the effectiveness of vaccination as a control strategy was analyzed in [31]. The role of immunization in reducing HBV transmission was further explored in [32], and a multi-group modeling framework was presented in [33]. Additionally, cost-effectiveness analyses of HBV interventions were conducted in [34], and the critical role of vaccination in disease control was mathematically assessed in [35]. Other papers included diffusion effects and time delays to further characterize persistence of infection and control strategies [36]. Other research has considered alternative aspects of HBV transmission and control [37], the impact of migration on disease dynamics [38] and different multi-intervention strategies [39]. These classical models, often with integer-order derivatives, led to a better understanding of vaccination dynamics, vertical transmission as well as disease dynamics within age or behavioral categories. But integer-order systems are not suitable for modeling memory and hereditary properties contained in biological process. To address this limitation, recent models based on fractional-order derivatives have been proposed to more accurately reflect HBV dynamics, especially for sexual transmission and regional heterogeneity. Together, these studies provide valuable insights for designing region-specific and behavior-informed public health interventions. Fractional calculus (FC) has emerged as a powerful tool in applied mathematics for addressing complex real-world problems. Fractional-order (FO) models are particularly advantageous over integer-order models due to their ability to capture memory effects and hereditary properties inherent in various systems [40–42]. Unlike integer-order derivatives, which are limited to describing instantaneous rates of change, fractional derivatives provide a more comprehensive analysis of the dynamics between two points in real-world phenomena. To overcome the constraints of classical local differentiation, several non-local FO differentiation approaches have been proposed in the literature. The Caputo [43], Caputo-Fabrizio [44, 45] and ADAtangana–Baleanu (AB) [45] fractional derivatives among them have received wide interests from scientists. The Atangana–Baleanu derivative in particular is well known for featuring a Mittag-Leffler (ML) memory kernel, as well as superior crossover representation of mean-square displacement, where it seems to be designed perfectly for a multitude of applications. Differentiation operators have applied been in a variety of engineering and physics systems. There applications have been successfully extended to biological and chemical models such as that of HIV/AIDS, HPV, covid-19, toxoplasmosis, tuberculosis, enzyme kinetics as well as pandemics responses [46, 47]. The flexibility and robustness of fractional-order models further enhance their suitability for capturing complex dynamics across diverse scientific domains [48, 49]. Using the AB fractional operator, researchers examined an HBV model that included a hospitalized class in [50] and an HBV model with treatment strategies in [51]. The Caputo derivative was utilized to study an HBV model incorporating hospitalization in [52]. Additionally, an FO HBV model that included the dynamics of antibody immune response was presented in [53]. Atangana and Baleanu recently made a critical commitment to fractional calculus by presenting operators for fractional integrals and derivatives in view of the Mittag-Leffler function [54]. The Mittag-Leffler capability has been demonstrated to be more successful than the traditional power function in describing problems related to natural phenomena. A key novelty of this study lies in the integration of HBV-induced disability into a fractional-order transmission model that accounts for both heterosexual and homosexual dynamics. This work introduces gender-specific disability probabilities and onset rates directly into the existing structure. This allows the model to reflect long-term functional impairments without increasing structural complexity. Moreover, by employing the Atangana-Baleanu-Caputo fractional derivative, the model captures memory and hereditary effects more effectively than classical integer-order models. Together, these innovations offer a more biologically and clinically realistic representation of HBV transmission, shedding light on the hidden burden of disability and guiding more targeted public health interventions. The remainder of this paper is organized as follows. In Sect. 2, we introduce the necessary preliminaries on fractional calculus operators. Section 3 presents the classical and extended HBV transmission models, incorporating heterosexual and homosexual transmission dynamics along with the disability burden. In Sect. 4, we derive the basic reproduction number R0 and analyze the disease-free equilibrium. Section 5 is dedicated to the existence and uniqueness analysis using fixed-point theory. In Sect. 6, we establish Hyers-Ulam stability of the model. Section 7 includes sensitivity analysis to determine the influence of key parameters on disease dynamics. Section 8 presents numerical simulations and discussion of the model’s behavior under various parameter settings. Finally, conclusions and public health implications are discussed in the closing section. Preliminaries This section gives an outline of fractional differential and integral operators, beginning with the Caputo fractional derivative. Caputo fractional derivative The Caputo fractional derivative is a broadly involved operator for fractional differentiation. First presented by Michele Caputo in 1967 [55], it has become fundamental in solving differential equations where the initial conditions are determined as far as integer-order derivatives. Unlike the Riemann-Liouville fractional derivative, this approach doesn’t need redefinition of fractional-order initial conditions. Mathematically, the Caputo fractional derivative is defined as: where , , and is the Gamma function [55]. An alternative representation is given as: This operator displays the property of vanishing when applied to constant functions. Besides, it empowers the utilization of Laplace transforms with initial values communicated regarding standard derivatives. For cases including a distributed fractional order, the Caputo derivative is formulated as: where is a weighting function. Expanding further: This formulation incorporates memory effects through the weighting function [55]. Caputo-fabrizio fractional derivative The Caputo-Fabrizio fractional derivative was introduced by Caputo and Fabrizio in 2015 [56]. It uses a non-singular kernel to refine fractional differentiation for functions f(t) of class C1. The operator is expressed as: where a < 0 and . Atangana-baleanu fractional derivative Atangana and Baleanu proposed a fractional derivative based on the generalised Mittag-Leffler function in 2016 [54]. Their aim was to introduce non-singular and non-local fractional differential operators. The Atangana-Baleanu derivative has both Riemann–Liouville and Caputo sense formulations. For a C1 function f(t), the Riemann–Liouville sense is given by [54, 57]: If the function f(t) is continuous, the Caputo sense derivative is represented as: The kernel of the Atangana-Baleanu derivative possesses characteristics of a cumulative distribution function. Specifically, the function rises along the real line and converges for every . The mapping forms a probability measure’s cumulative distribution function over positive real values. Consequently, this leads to the construction of an Λ-order Mittag-Leffler distribution. Notably, all such distributions are continuous. For , the Mittag-Leffler distribution reduces to the exponential distribution. The Laplace transform for these distributions is given by: This result indicates that the expectation diverges for , classifying these as geometrically stable distributions [54, 57]. Model formulation The classical HBV transmission model for the adult population, stratified by gender and infection status, is adopted from [58]. This baseline model captures transitions among susceptible, infected, carrier, and recovered individuals through both heterosexual and homosexual contacts. The model’s flow diagram is shown in Fig. 1. To extend the biological realism of this framework, particularly to account for the long-term health effects of chronic hepatitis B, we introduce a fractional-order version of the model that incorporates disability dynamics among carriers. The modifications are twofold: Inclusion of disability-associated parameters within the carrier compartments to reflect the onset of HBV-induced functional impairment; Inclusion of disability-associated parameters within the carrier compartments to reflect the onset of HBV-induced functional impairment; Reformulation of the system using the Atangana–Baleanu–Caputo (ABC) fractional derivative to model memory and hereditary properties inherent in HBV progression. Reformulation of the system using the Atangana–Baleanu–Caputo (ABC) fractional derivative to model memory and hereditary properties inherent in HBV progression. Flowchart of HBV transmission among adults, illustrating infection, progression, and recovery pathways in female and male compartments Disability resulting from chronic HBV infection is often observed in conjunction with cirrhosis and hepatocellular carcinoma and consists of a continuous reduction in physical function and cognitive ability. To represent these effects we have defined the parameters θf and θm, which are the fractions of male and female carriers respectively who are afflicted by disabling conditions. The rate at which the disabilities become clinically apparent is indicated by δ. The resulting fractional-order model, formulated in the ABC sense, is given as follows: The initial conditions are defined as: This formulation preserves the epidemiological structure of the classical model [58] while introducing fractional-order dynamics and disability-specific transitions, offering a more comprehensive representation of HBV progression and its long-term health consequences. The model partitions the total adult population into seven dynamic compartments that represent distinct infection and health states. Female and male populations are treated separately to capture gender-specific transmission patterns and disease progression: and : susceptible females and males who are at risk of acquiring HBV infection; and : susceptible females and males who are at risk of acquiring HBV infection; and : acutely infected females and males capable of transmitting the virus; and : acutely infected females and males capable of transmitting the virus; and : chronically infected female and male carriers who remain infectious and may develop HBV-related disability; and : chronically infected female and male carriers who remain infectious and may develop HBV-related disability; : recovered individuals who have acquired long-term immunity. : recovered individuals who have acquired long-term immunity. Descriptions of the model parameters and their assigned values are presented in Table 1. Table 1Model parameters, assigned values, and their epidemiological interpretationParameterValueDescriptionSourcea0.75Proportion of female susceptibles engaged in heterosexual contact [58]b0.75Proportion of male susceptibles engaged in heterosexual contact [58]AfAnnual recruitment rate of females into the susceptible class [9]AmAnnual recruitment rate of males into the susceptible class [9]µ0Natural mortality rate [58]µ1Additional mortality rate associated with HBV infection [58]0.16Scaling factor for reduced infectiousness of chronic carriers [12]γ1Rate at which acute infections progress to the carrier stage [58]γ2Recovery rate from the carrier stage to the immune state [12]γ3Adult vaccination rate [9]qf0.05Probability that an infected female becomes a chronic carrier [59]qm0.07Probability that an infected male becomes a chronic carrier [59]βTransmission coefficient for non-sexual exposure routes [58]βmfTransmission probability per male-to-female sexual contact [58]βfmTransmission probability per female-to-male sexual contact [58]βff0.0142Transmission probability per female-to-female sexual contact [58]βmmTransmission probability per male-to-male sexual contact [58]cmfAverage annual number of male-female sexual contacts (female perspective) [58]cfmAverage annual number of female-male sexual contacts (male perspective) [58]cffAverage annual number of female-female sexual contacts [58]cmmAverage annual number of male-male sexual contacts [58]θf0.10–0.30Fraction of female carriers developing chronic HBV-related disabilityAssumedθm0.15–0.35Fraction of male carriers developing HBV-related disabilityAssumedδ0.05–0.10 Average rate of onset of HBV-induced disability (10–20 year latency)Assumed Model parameters, assigned values, and their epidemiological interpretation For simplicity, let us define Deterministic basic reproduction number The model (9) has a disease-free equilibrium given by: where: We calculate the basic reproduction number R0 using the methods in [60, 61]. Let: where: The disease-free equilibrium is: The Jacobian matrix at is given by: where H and J are matrices, and: The matrix V is given by: where: Additionally: Following [62, 63], the basic reproduction number R0 is the spectral radius of the matrix FV−1, denoted as . After calculation: where: and: Existence results Assume that B(J) represent a real-valued continuous function which containing the suprimum norm space property is a banach space on and with norm , where . Now by setting fractional integral operator to left and right hand sides of Eq.(9), we get the following system Applying the Atangana-Baleanu-Caputo fractional integral operator to the both sides of Eq.(9), we obtain where If and Cm contains their upper greatest value or bound, then , , , , , and must fulfill the Lipschitz condition. Assuming that Sf and are two different functions, we obtain Taking into account we obtain the inequality In a similar way, we can get the following The last equation is the Lipschitzian condition that has held for all the mappings. Going in a repetition mode, the equation in (14) becomes with initial conditions: , and Upon applying the recursive update definitions, we obtain: It is important to note that: Furthermore, based on the results of (17)–(18) and defining the differences as: we arrive at Eq. (21). Theorem 5.1 Note that the given conditions holds Then, (9) has unique root for . Proof This is derived that , , and , are mapping which have bounds. Next, as one can see from Eqs. (17) and (18), the abbreviation , , and holds for Lipchitzian condition. Hence, applying Eq. (21) along with a repeating hypothesis, we get as Therefore, it implies for , all mapping exists and fulfill Furthermore, from Eq. (23) and applying the triangle in-equality, for any k, we have with by supposition. As, , and Rn may be observed as a Cauchy sequence in banach space B(J). It implies that all the quantities are uniformly convergent [64]. applying the limiting Theorem in Eq. (20) as conforms that the limit of such type of sequences have unique root of (9). This shows that the existence of Eq. (9) is unique having the condition (22). Hyers-Ulam stability Definition 6.1 The AB arbitrary order integration problem as in Eq. (14) is called H-U stable [62] if their exist fixed fulfilling: For all , for there exist which are satisfying Such that Theorem 6.1 Under the condition J, the considered model of arbitrary order (9) is H-U stable. Proof By Theorem (5.1), the proposed AB fractional problem (9) has unique root fulfilling equations of model (14). Then as follows Taking, this implies Similarly, we have the followings So the derivation is achieved. Sensitivity analysis This section provides a detailed examination of the model through local and global sensitivity analyses to assess parameter impact on the transmission dynamics of HBV. Local sensitivity analysis Sensitivity analysis is an important tool for identifying parameters that most strongly influence disease transmission. In epidemiological modeling, it helps determine how small perturbations in model parameters affect key outcomes such as the basic reproduction number (R0). Here, we apply the local (forward) sensitivity approach, which measures the normalized rate of change of R0 with respect to each parameter. Definition 7.1 The normalized forward sensitivity index of R0 with respect to a differentiable parameter is defined as This is indicated by the analysis presented in Table 2 and Fig. 2. It is clear that the parameters βmm, Af, Am, cmm, βff, and cff exhibit very high positive sensitivities, suggesting that recruitment and contact are both important means by which HBV can spread to a greater degree. In contrast, the parameters γ1, b, and γ3 show large negative sensitivity values, indicating that better transmission control, faster recovery from infection or vaccination, will all bring about a substantial reduction in R0. The disability related parameters (θf, θm, and δ) demonstrate smaller but notable effects, reflecting that it is not so much their modification of the infectious period itself that is critical here, but rather the impact of those parameters on the dynamics of the population. The graphical results showing the effect of selected parameters on the basic reproduction number R0 are displayed in Fig. 3. Table 2Normalized forward sensitivity indices of the R0 with respect to model parameters. Positive values indicate parameters that increase R0, while negative values indicate those that reduce itParameterSIndexValueParameterSIndexValueaSa−0.71338bSb−1.5007AfSAf0.36726AmSAm0.63274βSβ0.26193γ1−1.3793γ2−0.41326γ3−1.3085µ0−0.33313µ1−0.033061qfSqf0.021231qmSqm0.0500530.071284βff0.23779βmm0.50025βmf3.31e-09βfm2.05e-05cmfScmf3.31e-09cfmScfm2.05e-05cffScff0.23779cmmScmm0.50025θf−0.24917θm0.27142δSδ0.022253 Normalized forward sensitivity indices of the R0 with respect to model parameters. Positive values indicate parameters that increase R0, while negative values indicate those that reduce it Parametric sensitivity analysis of the basic reproduction number (R0). Positive and negative bars indicate the parameters that increase or decrease R0, respectively, highlighting their relative influence on disease transmission dynamics Contour plots showing the combined effects of key parameter pairs on the R0. Subfigures (a–f) illustrate the interactions between , , , , , and . The contour gradients show how simultaneous variations in these parameters influence R0, revealing nonlinear dependencies and potential synergistic or antagonistic relationships Global sensitivity analysis (LHS–PRCC) In addition to the local (forward) sensitivity analysis, a global sensitivity analysis (GSA)was performed in order to consider joint and nonlinear effects among various parameters on R0. The Latin Hypercube Sampling (LHS) technique was applied to perform a global analysis in combination with Partial Rank Correlation Coefficients (PRCC), a popular method in epidemiological modeling for assessing the importance of uncertain parameters on the model predictions. This method is effective for searching the high-dimensional parameter space and depicts nonlinear and interaction effects. A set of 1000 parameters were created, where each parameter was fluctuated around 20% of its baseline value in order to sufficiently span the parameter space. Pairwise rank correlations between each input parameters and the resultant R0 values were subsequently computed. The larger is the absolute value of these PRCCs, the more a parameter influences R0, and its sign allows to tell if that parameter increases (for positive values) or decreases (−ve values) the reproduction number. The global analysis identified six parameters as being most influential: the vaccination rate (γ3), male-to-male transmission probability (βmm), female-to-female transmission probability (βff), male disability onset rate (θm), female disability onset rate (θf), and disability onset speed (δ). Among these, γ3 showed a strong negative correlation (), confirming its significant suppressive effect on HBV transmission. Conversely, βmm (PRCC = 0.957) and βff (PRCC = 0.838) exhibited strong positive correlations, highlighting their dominant contribution to increasing R0. The disability-related parameters θm, θf, and δ demonstrated moderate influence, reflecting the role of disease progression and disability in shaping transmission potential. Figures 4, 5, and 6 summarize the results of the global analysis. The PRCC bar graph is presented in Fig. 4, whereby colored bars represent the direction and magnitude of parameter influence. In Fig. 5 the most influential six parameters in the model give histograms showing the uniformity of sampling, which has been established within their ranges. In Fig. 6 scatter plots of R0 against each of the key parameters are shown, which helps to illustrate the positive and negative relationship trends which were calculated using the PRCC values. Overall, the results of the global sensitivity analysis substantiate findings obtained from the local sensitivity analysis. The parameters γ3, βmm, and βff remain the most important drivers of the HBV transmission dynamics. Inclusion of disability parameters (, and δ) in the modeled biological reality adds further biological validity and emphasizes how disability changes subtly the effective infectious period. In summary, these results give a much more comprehensive understanding of the response of the model to parameter variation and reveals the important role played by vaccination and behavioral interventions in reducing R0. Partial rank correlation coefficients (PRCC) showing the global sensitivity of R0 to all model parameters Histograms obtained from Latin hypercube sampling using a sample size of 1000 for the six input parameters Scatter plots for the basic reproduction number and six sampled input parameters values. These results were obtained from Latin hypercube sampling using a sample size of 1000 Iterative solution by Newton Polynomial In this section, we present numerical methods for the proposed fractional order model, using Newton polynomial interpolation [63]. Recently, Atangana and Seda applied the scheme to a new COVID-19 model [65], which were solved successfully using Newton polynomials. Newton interpolation is a classical model for polynomial interpolation that is of great importance in many fields, such as numerical analysis and image processing. Standard interpolation techniques describe a particular designed function for a given dataset. Newton’s method of polynomial interpolation, on the other hand, presents more advantages than other methods. Some of these include fast convergence, ease of implementation, and less complexity in computations including integration and differentiation. Moreover, it is also possible to evaluate derivatives of any order in this scheme quite simply. By adjusting the parameters, the shape of interpolation curves or surfaces can be conveniently altered within the scope of geometric design by varying the Newton-type polynomial interpolant function parameters within the interpolation domain. We commence the discussion with the Mittag-Leffler kernel: More simply, we can write as follows; Later on after application of fractional integration with Mittag-Leffler kernel law and plugging Newton polynomial in type of equations, we evaluate our model as follows; Where Simulation results and discussion In this section, we introduce a numerical scheme for solving the proposed model, employing the Newton polynomial interpolation method [65]. To explore how key parameters influence the spread of HBV, we numerically simulate the model given in (9). The simulations are performed over a period of 50 months, and in some cases extended to 100 months to assess the system’s long-term dynamics and stability. To monitor infection trends among adult males and females, we introduce the following normalized state variables: The corresponding initial values are: The total initial population is set to . These initial conditions were optimized in MATLAB using data reported by the CDC [58]. The numerical values of all parameters used in the simulations are listed in Table 1, and the resulting simulation outputs are illustrated in Figs. 7, 8, 9, and 10. Simulation profile of each state variable of model (Eq. 9) using the Mittag-Leffler Generalized Function for time t = 50 months at the different fractional-orders Simulation of each state variable of the fractional-order epidemic model (Eq. 9) using the Mittag-Leffler Generalized Function for t = 100 months at fractional orders There dimensional simulation profile of each state variable of the model (Eq. 9) over time (t) and fractional-order (Ψ1), using the Mittag-Leffler generalized function. These simulation results exhibit the consolidated impact of memory impacts (Ψ1) and time on the population dynamics Simulation of the recovered population R(t) for different fractional orders Ψ1, based on Eq. (11). (a) shows dynamics over 50 months, and (b) over 100 months. Results highlight the influence of memory effects on recovery using the Mittag-Leffler kernel The fractional-order epidemic model in the Atangana-Baleanu-Caputo sense provides insightful simulations as depicted in Fig. 8a-f, where the influence of varying fractional orders (Ψ1) on disease dynamics is analyzed. The simulation results reveal that the female susceptible population () decreases at different rates depending on the fractional order (Ψ1). For lower fractional orders (e.g., ), the memory effects slow down the transition of individuals to the infective or recovered states, resulting in a slower decline in . Conversely, higher fractional orders () depict a sharper decrease, reflecting reduced memory effects and a more rapid disease spread. Similarly, the female infective population () peaks at different times depending on Ψ1, with lower fractional orders resulting in delayed peaks due to prolonged infection periods caused by stronger memory effects. While, the higher orders lead to earlier peaks, representing faster transitions towards the recovery or chronic states. The () class carrier population grows steadily for all values of the fractional order Ψ1, with lower orders sustaining a higher carrier individuals over time, prediction of prolonged persistence of carriers due to memory effects. For the higher fractional orders showes a slower accumulation of carriers, reflecting faster recovery dynamics. Furthermore, the dynamics () of the male susceptible population similar behaviour those of , where lower fractional orders show a slower depletion of susceptible males, emphasizing the influence of memory effects in delaying the progression from susceptibility to recovery or infectivity stages. For higher fractional order Ψ1 values, the faster decreases of corresponds to a more faster spread of infection among the male individuals. The class () male infected population showes significant dependence on the fractional order. Lower values of Ψ1 lead to slower infection process due to greater memory effects in the system. As can be seen from the simulation if Ψ1 increases, then the male infected population increases quicker signifying fast transistions from the susceptible state to the infected stage. Finally, the male transporter population () denotes patterns similar to the female carrier population where lower fractions lead to a larger transporter population eventually, whereas higer value of effects less persistence of carriers due to increased speed of transition to recovery or lower infectivity. The fractional order Ψ1 controls the memory effects and non-local interactions in the model. Lower values of fractional order deal with frameworks with more considerable memory in which the individuals of the system retain past states for greater time, so that the transition between compartments is slower. Higher values of fractional order tend towards classical model of integer order where dynamics are faster and effects of memory minimized. This technique is representative of realistic epidemic behavior, it also advances a versatile tool in the analysis of complex disease systems. The recovered population () huge reliance on the on the fractional order, with slower recovery processes for lower Ψ1 values due to stronger memory effects. As Ψ1 increases, the recovered population grows more rapidly, indicating faster transitions from the infective and carrier states to recovery. The 3D plots reflect the time changes, and force orders effects on each of the population compartments as in Fig. 9a-9f. The susceptible female population decreases differently to a particular level with respect to the fractional order . For small , the memory effects inhibit transitions such that susceptible vanish more slowly. On the other hand, for larger fractional orders depletion is faster indicating lesser memory effects and rapid disease progression. For the , female infective, peak ages are clearly present which is also determined by Ψ1, with smaller fractional orders also exhibiting delayed and persistent peaks resulting from longer life of the disease. On the other hand, higher fractional orders accelerate transitions, leading to earlier and sharper peaks. The () carrier population of model grows steadily over time, with lower Ψ1 values allowing carriers to persist longer, while higher Ψ1 values result in faster recovery dynamics and reduced carrier persistence. The male susceptible class exhibits the same patterns, showing slower depletion under lower fractional orders and quicker transitions under higher orders. The male infective class shows corresponding delayed peaks under lower values of Ψ1, indicating prolonged infective states due to strong memory effects. The male carrier class shows close conformity with the female carrier trends, here strong memory effects (lower Ψ1) sustains the carrier class, and weak memory (higher Ψ1) causes a decline in the period of the carrier state and marked it easier to attain recovery. These simulations printout the importance of fractional order as a significant parameter which effects memory effects and system dynamics. Lesser orders of fractionalize represents systems which possess strong memory and in which past performances strongly influence performance, and consequently in which transitions between these compartments is slower than in high fractional order systems. The higher fractional orders approximate the classical systems more closely, in which the dynamics are more localized, and so transition is effected more readily. The employment of this fractional system does imply an ability to more correctly represent the complexities of the epidemic behaviours, and provides a good theoretical model for the understanding and control of disease dynamics under varying memory conditions. The recovered population of the model depends heavily on Ψ1, with lower fractional orders slowing recovery and higher orders accelerating it, resulting in quicker transitions to the recovered state. Conclusion This study develops a fractional-order mathematical model to describe the heterosexual and homosexual transmission dynamics of HBV, extending existing frameworks to include HBV-induced disability. By applying fractional calculus with the Mittag-Leffler kernel, the model captures the memory and hereditary characteristics inherent in biological systems. Analytical results concerning the existence, uniqueness, and Hyers–Ulam stability of the model confirm its theoretical soundness. A numerical method based on Newton polynomial was employed to simulate the dynamics of the system over time with a sufficient degree of precision. One of the main contributions of this study is the introduction of the disability parameters , and δ, which describe the incidence and manner of chronic HBV carrier long-term functional disability, while preserving the original compartmental structure. This contributes to greater biological realism and more thorough evaluation of the overall burden of disease. Sensitivity analysis of the basic reproduction number (R0) shows that the major driving forces in the spread of infection are the rates of recruitment (), transmission probabilities (βmm , βff ), and the contact rates (). Parameters which appear to lower R0 significantly are the vaccination rate (γ3), the disease progression rate (), and the disability related characteristics, especially the parameters θf and δ. The parameter θm has a slight positive effect, probably because of behavioral differences in male carriers before disability onset. Graphical results bring out some of the interrelationships in the parameters, and point to the need for integrated approaches to intervention. The inclusion of the disability dynamics also indicates an additional load of epidemiological disease and emphasizes the fact that functional health results are necessary for a complete understanding of HBV transmission. Acknowledgements The authors extend their appreciation to the King Salman Center for Disability Research for funding this work through Research Group no. KSRG-2024–184. Abbreviations Hepatitis B Virus Hepatitis B surface antigen World Health Organization China–GAVI Project Men who have sex with men Women who have sex with women Hepatitis E Virus Hepatocellular Carcinoma Disability-Adjusted Life Year Years Lived with Disability Years of Life Lost Body Mass Index Susceptible–Exposed–Infectious–Recovered Ordinary Differential Equation Fractional Differential Equation Atangana–Baleanu–Caputo Mittag-Leffler Fractional Calculus Fractional Order Caputo–Prabhakar Riemann–Liouville Author contributions K.G., R.Z., B.M.M., and H.A.N. conceptualization and validation of the model development. B.M.F. and O.J.P. developed the mathematical model, performed analytical and numerical simulations, and interpretation of the results. O.J.P.,K.G., R.Z. drafted the manuscript, while all authors contributed to reviewing and editing the final version. All authors read and approved the final manuscript. Funding The authors extend their appreciation to the King Salman Center for Disability Research for funding thiswork through Research Group no. KSRG-2024–184. Data availability All data generated or analyzed during this study are included in the paper. Declarations Ethics approval and consent to participate N/A. Consent for publication N/A. Conflict of interest The authors declare no conflict of interest. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement All data generated or analyzed during this study are included in the paper.