Abstract
Households (HH) have been traditionally described as the main environments where people are at risk of dengue (and other arbovirus) infection. Mounting entomological evidence has suggested a larger role of environments other than HH in transmission. Recently, an agent-based model estimated that over half of infections occur in non-household (NH) environments like workplaces, markets, and recreational sites. However, the importance of human and vector mobility and the configurations of urban spaces in mediating the effects of NH on dengue transmission remains understudied. To improve our knowledge of the relevance of these spaces in transmission, we expanded an agent-based model calibrated from field data in Kenya to examine movement of people and vectors under different spatial configurations of buildings. In this model, we assessed the number of people traveling between HH and NH, their distance, and the number of mosquitoes migrating. Those were studied on three different urban configurations, on which the NH are spatially distributed either randomly (scattered), centered (in a single center), or clustered (in more than one cluster). Across simulations, the number of people moving is a major influential variable where higher levels of movement between HH and NH increases the number of cases. In addition, the number of cases is higher when NH are scattered. These results highlight the importance of NH in transmission as a major spreader of infections between households. In addition, we found an inverse relationship between mosquito migration and population size and hence transmission, which underscores the high cost of migration for mosquitoes. Intriguingly, the distance that people travel from HH to NH seems to have little effect on dengue burden; however, it affects the level of spatial clustering of cases. The results of this work support the relevance of NH in transmission and its interaction with human movement in driving dengue dynamics.
Author summary Recent evidence describes a major role of non-household (NH) environments in dengue transmission. This new knowledge implies that spatial distribution of these locations and the movement of both vectors and humans among them has implications in dynamics of transmission. However, these haven’t been evaluated neither together nor considering a differential role in transmission of urban environments. We modified a previously informed agent-based model to include spatial variables by assessing three different urban conformations, i.e. when NH are randomly distributed (scattered), centered, or grouped in differed clusters (clustered). On these, we also evaluated the movement of both vectors and people from households (HH) to NH (both distance and number of individuals). Our findings, which includes a higher burden of dengue when NH are scattered and at higher levels of human movement and low level of vector migration, are aligned with the idea that infections are mainly happening in NH mediated by human mobilization. Infected people reach the households where local subpopulation of vector spread the virus to remaining inhabitants. This work highlights the role of NH and human mobility in dengue transmission.
Introduction
Dengue is a vector-borne disease prevalent and on the rise in most of the tropical and subtropical regions around the globe. The main vector, Aedes aegypti, is highly anthropophilic, found very close to human environments and impacting the public health of urban environments [1, 2].
Historically, the main strategy to control the disease has been to reduce vector-human contact by reducing the size of mosquito populations. Conventional wisdom is that for effective control, these activities should be focused on households (HH) as the main environment where transmission is happening [3, 4]. However, some studies have suggested that locations other than households might have an important role because of a significant presence of mosquitoes [5–8], and infected vectors [9]. Recently, a study using agent-based modeling to quantify the number of infections in different types of urban spaces estimated that over half of infections are happening in non-household (NH) environments, where the main high-risk spaces are workplaces and markets/shops [10].
These results have implications for dengue epidemiology since the high influx of people in NH suggests that these spaces can contribute to the spread of infections. In this way, the total number of infections can be affected by the distribution in space of NH and the movement of people between HH and NH. However, these intra-urban dynamics have never been described in the context of different roles among HH and different NH space types [11–14], though Massaro and colleagues used mobile phone data to get estimates for movement between workplaces [15].
Building on a previous result showing the major importance of NH for dengue transmission, we now evaluate how levels of human and vector mobility as well as different urban spatial configurations of NH affect dengue transmission. In particular, what role does spatial configuration of NH spaces play, along with the extent to which people and mosquitoes move between spaces, in determining dengue dynamics? To address this, we modified a previously reported agent-based model [10] to make it spatially explicit by assigning coordinates that mimic different urban conformations and evaluated different scenarios of movement of people and vectors. We then assessed how these variables affect the burden of dengue and the spatial patterns to understand urban-level transmission dynamics.
Methods
Model overview
To achieve the aims of this study, we modified the agent-based model previously used to describe the importance of HH and NH in transmission [10]. The model was developed to quantify the relative contribution of five different types of NH (workplaces, markets or shops, recreational, religious, and schools) and HH to dengue burden. The model development and calibration were based on data from the two Kenyan cities of Kisumu and Ukunda [10]. This paper’s results focus on parameters calibrated to Kisumu, although additional results including dynamics from Ukunda are found in supporting information.
The model represents the movement of people to daily-commuting locations like schools and workplaces as well as other locations, for which both the number and identity of people who visit them is randomly defined every day. The latter include spaces like markets or shops, recreational and religious spaces. Movement between HH is also included in the model. Based on vector surveys conducted over two years of fieldwork previously published [16], NH and HH environments were assigned to have mosquito presence or absence based on observed prevalence of mosquitoes. Population dynamics of vectors were modeled at the building level, whereby the sub-population dynamics are determined based on building-level conditions like the presence of water containers and the total amount of water they can hold, again informed by field vector surveillance data. These dynamics as well as infection dynamics of vectors are also determined by temperature by using functions previously described and widely used elsewhere [17–19].
Transmission events happen in those locations where infected vectors contact susceptible humans or vice versa. Mosquitoes bite depending on both temperature-dependent biting rate and the probability of having a successful vector-human encounter, which depends on the amount of time that humans spend in the location. Infection status of mosquitoes can be susceptible, exposed, or infected while humans can be either susceptible, exposed, infected, or recovered and (temporarily) immune. Since the model does not explicitly represent dynamics of different serotypes, recovered humans go back to susceptible status after three months in recovered status. The number of infections in each location is recorded daily. Statistics about the total number of infections and locations are provided weekly. The model simulates transmission dynamics happening for 731 days (comprising temperature conditions between January 1st of 2020 until December 31st of 2021) and results are shown as a distribution of the number of infections over 200 simulations. The model and modifications described in this work were coded in Julia language (v1.10.0) and simulations were run on Sherlock computational cluster (Stanford Research Computing Center).
Spatial variables
To include mobility-associated variables, we made the model spatially explicit. No actual spatial coordinates were used so we could record the outcome when different urban conformations are tested by using the same set of spatial coordinates. In this way, we can assure that differences are due to the building designation as HH or NH and not by the specific spatial location. Because the model considers populations of about 20,000 people, the total municipality areas were rescaled to fit the total number of structures of the virtual populations while considering similar densities. Spatial coordinates were randomly generated and assigned to each structure of the population. Coordinates assignments were done based on either “Scattered” (randomly distributed), “Centered” (majority of NH concentrated in the center of the city in a single cluster), or “Clustered” (majority of NH concentrated in three clusters) configurations (Fig 1), so no new coordinates were created for each spatial distribution, but spatial assignments of building types to locations differed among them.
scattered (NH randomly distributed in space), centered (majority of NH are clustered in the center), and clustered (majority of NH are grouped in three clusters). Only NH are shown while HH are interspersed among NH. Religious and schools are represented in the first row, markets/shops and recreational in middle row and more frequent NH, workplaces, are represented in the lower row.
Movement of people
We included two movement-related variables: distance from HH to NH locations, and number of people moving. To control the distance of people moving, we limited the attendance of people to (1) those NH locations that are the closest to their respective HH (categorized as distance zero, 0), (2) those locations that are at least 500 meters away from the respective HH, or (3) locations at least 1000 meters away from the HH. The three treatments were applied to each of the three urban conformations. However, on the clustered or centered conformations, the closest distance treatment would favor NH locations at the periphery of the cluster. To avoid this, as many close NH were assigned to each HH as inhabitants it had.
The number of people visiting NH was simulated at three levels. First, we included the same levels previously described in the model, categorized as 100% mobility [10]. This treatment includes all students and workers attending their respective school and workplace, and random-attendance locations (religious, markets/shops, and recreational) with number of people uniformly distributed with parameters minimum = 10 and maximum = 70. The number of people visiting a given location is determined daily. We also decreased the number of people moving to school and workplaces to 50% and remaining NH locations to a uniform distribution with parameters minimum = 5 and maximum = 35. Finally, we include a level of people movement of 20% for school and workplaces and uniform distribution with parameters minimum = 2 and maximum = 14 for random-attendance locations.
Movement of mosquitoes
We consider two variables to estimate a baseline migration probability for the number of mosquitoes moving from given locations: availability of both breeding and blood-feeding resources.
The model includes a local density-dependent function that allows for the mosquito population to grow in a location-specific way, allowing it to represent dynamics previously described for a fragmented environment [20]. By using this function, the number of mosquitoes in each location grows according to availability of water containers, the density of immature mosquitoes in those containers, and temperature (peaking at 29°C), as follows:
Where
In the equation, D is the larval density expressed as the ratio of the number of larvae to liters of water available for breeding in the structure, and d(T) is the term describing the temperature-dependence of population growth. The equation allows the sub-population to grow when larvae density is low by assuming they occupy all the water containers inside or around the structure. As the mosquito sub-population grows, resource availability in the water declines, slowing population growth. By considering Nmtas the mosquito subpopulation size at time t, when the ratio Nmt/Nmt+1approaches 1, breeding site resources are reaching their carrying capacity and hence, the probability of a mosquito migrating increases. Because there is a strong stochastic component in estimating Nmt+1, by running 20 simulations for every location every day, we estimated a parameter MP as the median of ratios Nmt/Nmt+1.
On the other hand, we also consider the probability that each female mosquito bites a human within a given location. The rate of biting is temperature dependent (a), according to previously published [17]. We coupled it into a binomial distribution to define the number of biting mosquitoes as follows:
Where Nb is the number of females needing to bite on day t and Nmt is the size of sub-population mosquitoes that day. While most models assume that all females bite when they need to, here we consider the possibility that mosquitoes cannot find blood meal hosts in a given place and time as dependent on the human occupancy of a given location. We assume that there is a probability for Nb mosquitoes to have a successful biting human-mosquito encounter (P(bite)) based on the size of Nb and the amount of time that humans spend in the location, as follows:
Where Nht is the number of people attending a given location on day t for a structure-specific number of hours. Finally, the actual number females that fed (NF) in each day, is estimated by
In this sense, when the number of people on a given day is low and the number of hours they spend in the location are few, the probability of a female mosquito successfully biting is low. So, as the ratio NF/Nb(later called MF) approaches 0, the probability of mosquitoes migrating increases. Again, for every location every day, 20 simulations were run to estimate a distribution of MF.
Finally, we estimated the migration probability as the product Mp·(1-MF), which depends on both larval density conditions via Mp and adult biting conditions via MF. The resulting probability was used in a binomial distribution to determine the number of migrating mosquitoes each day. To evaluate the effect of different mosquito movement regimes on the burden of dengue, we considered migration rates of 100% (as unmodified product Mp·(1-MF)), 50%, and 10%. The total number of cases was quantified after 200 simulations.
Once a mosquito migrates, a new location is assigned by considering a dispersal kernel [21]. Following previous work [22], we used a lognormal function with the form
Where d is distance and both a and b are parameters to be estimated. We fitted a function by assuming a mean dispersal distance of 105.69 meters, as estimated for Aedes aegypti by [23].
Spatial autocorrelation
By taking advantage of spatial features introduced in the model, we performed a spatial autocorrelation analysis to evaluate the level of clustering of dengue cases recorded in each simulation as a function of urban configuration and human and mosquito mobility. To do this, the household location of each infection is recorded so a Global Moran’s I index could be estimated at the end of the simulated period. Global Moran’s I ranges from –1 to +1 where –1 means totally dispersed location of cases while a value of 1 represents a total spatially clustered distribution (total separation between locations with dengue cases and those without cases). In this sense, the null hypothesis of this analysis is that dengue cases are randomly distributed in municipalities, represented by Moran’s I value of 0 [24]. The analysis was done for every simulation by using 1,000 permutations for inference in each of them. Analyses were done using the package SpatialDependence.jl implemented in Julia language (v 1.10.0).
Results
Burden of dengue is strongly affected by number of people visiting NH
We quantified the total number of cases after two years of transmission. When we simulated the epidemic under different human movement regimes, it was evident that the number of cases decreased as the number of people moving from HH decreased. For 100% human movement, irrespective of the urban conformation, we estimated for a median of 4,228 cases (IQR: 3,025 – 4,921), which decreased to a median of 764 (IQR: 349 – 1,626) and 154 (IQR: 108 – 232) cases for 50% and 20% human movement, respectively (Fig 2 and S1 Table) (all results are derived from the model calibrated for Kisumu, Kenya; see supporting information for further results for the Kenyan city of Ukunda).
Three levels of human movement were assessed (20%, 50%, and 100%) on three urban conformations (scattered, centered, or clustered). Boxplots shows the distribution of the total number of infections for 200 runs of two-year simulation where median is the horizontal line, the filled box is the interquartile (IQR) range, the whiskers show the values above and under the IQR and no more than 1.5·IQR, and dots are representing values beyond this range.
Additionally, the scenario where NH locations are spatially randomly distributed produced more cases, though at all movement levels the interquartile ranges for different spatial configurations overlapped (Fig 2). Thus, scattered conformation yielded a median of 4,672 (IQR: 3,956 – 5,227) while the centered and clustered scenarios produced, respectively, medians of 4,432 (IQR: 3,587 – 5,027) and 3,178 (IQR: 1,785 –4,179) (S2 Table).
When we compared HH and NH environments, the number of infections is higher in NH at higher levels of movement. However, at lower levels of movement and hence lower transmission, the number of infections between HH and NH becomes roughly similar. In this way, at 100% of movement irrespective of urban conformation, NH produced 67% of the cases, but this proportion decreased to 58.8% and to 42.3% at 50% and 20% of human movement, respectively (Fig 3 and S1 Table).
Three levels of human movement were assessed (20%, 50%, and 100%) on three urban conformations (scattered, centered, or clustered). Boxplots are showing the distribution of total number of infections for 200 runs of two-year simulation where median is the horizontal line, the filled box is the interquartile (IQR) range, the whiskers show the values above and under the IQR and no more than 1.5·IQR, and dots are representing values beyond this range.
Distance from HH to NH makes little difference in dengue burden but defines level of spatial structure
Interestingly, when we evaluated the number of infections produced by different distance regimes from HH to NH, only slight differences were evident among them. These are mainly related to an increased number of cases when NH are clustered (Fig 4 and S3 Table). Besides these slight changes, differences among urban conformations are still evident (Fig 4).
Distance 0 means people only visit the closest NH location from HH. On the three urban configuration scenarios, three levels of distance from HH to NH were assessed (the closest [categorized as 0], at least 500 meters, and at least 1000 meters). Boxplots are showing the total number of infections after two-year simulation for 200 runs.
Intrigued by the apparent lower importance of human movement distance in transmission, we wanted to explore further by assessing the spatial structure of cases. Given that the number of people moving affects the number of cases, we evaluated the spatial structuring when distance traveled is considered as well. In general, we found that irrespective of urban conformation, when people move short distances the level of spatial structure is higher, as expected (Fig 5). Similarly, the spatial structuring levels are also modified by the number of people moving, where lower levels of movement decrease Moran’s I, thereby making cases more dispersed (Fig S6).
Distribution of Moran’s I values for 200 simulations for each of the three distance regimes is shown. Both significant (α = 0.05) and non-significant are displayed by color. Three levels of distance from HH to NH were assessed (the closest [categorized as 0], at least 500 meters, and at least 1000 meters) on three urban conformations (scattered, centered, or clustered).
Movement of mosquitoes is costly and reduces human dengue burden
When we considered different regimes of mosquito movement, the model yielded a lower number of cases when mosquito migration was higher. In this sense, the median number of cases was 306 (IQR: 222 – 419) at 100% of mosquito movement and increased to 1,090 (IQR: 533 – 1,714) and 4,670 (IQR: 4,005 – 5,229) when number of mosquitoes was 50% and 10%, respectively (S4 Table). Still the proportion of infections happening in HH compared to NH remained lower regardless of mosquito movement (Fig 6). To better understand the mechanism by which mosquito movement was reducing transmission, we studied the size of mosquito populations for each mosquito movement regime, which resulted in smaller population sizes at higher movement rate (Fig S9), indicating the importance of mosquito mortality during migration as a limit on transmission.
Three levels of vector movement permissiveness were assessed (100%, 50%, and 20%). Human mobility was kept to 100% for all simulations, so differences are strictly due to mosquito movement. Boxplots are showing the total number of infections after two-year simulation for 200 runs.
Discussion
Recently, results of this model have described the important role of NH environments for transmission, which is greater than previously recognized, accounting for over half of infections and implying an important role of vector control in NH spaces [10]. Here, we aimed to further understand how this differential role of NH environments interacts with movement and spatial configurations of urban environments to drive dengue transmission. We showed that human movement is a primary driver of dengue dynamics, irrespective of urban spatial configuration, and that mosquito movement tends to reduce transmission by increasing mosquito mortality. Further, qualitatively similar, outcomes generated by simulating conditions from the Kenyan city of Ukunda supports these results and can be found in supplemental material.
Among urban conformations, when NH spaces are scattered throughout the city it allows for closer connections to HH and therefore increased transmission. In this way, NH spaces serve as spreaders of infection since they are highly visited locations, which increases the chances of having a successful feeding encounter between humans and infected vectors. Once an individual is infected, the chances of infecting mosquitoes inside the household and in turn having another household inhabitant infected increases, generating local household chains of transmission.
For this reason, reducing the number of individuals visiting NH locations reduces the burden of dengue (Fig 2). In this sense, when number of people visiting NH decreases, the number of infections happening in these spaces is also decreased until becoming roughly even with number of infections in HH. These results are supported by previous reports that when COVID-19 pandemic lockdown forced people to stay at home most of the time, the number of dengue cases was significantly reduced [14, 25, 26]. The idea of human movement guiding different levels of transmission is not new and has been explored previously, especially by Stoddard and colleagues (2009) who found that human movement is even more important for transmission than vector abundance [11]. However, here we highlight how human mobility interacts with NH spaces as drivers of transmission, which then disseminates within households. As a result, when control is focused on households, it prevents the spread of disease to the remaining household inhabitants. However, broader focuses of transmission remain active as long as NH transmission is not under control.
Distance from HH to NH was not as important as overall levels of human movement for dengue burden (Fig 4). The lack of influence of this variable supports our previous result by suggesting that it is not how far people are traveling but the destination and total amount of movement. This is in line with previous work where a large, longitudinal study in Iquitos, Peru showed that human infection risk is mainly driven by individuals visiting locations with presence of infected vectors, irrespective of the distance [12].
It is important to note that our model does not account for movement times, which increase with distance. These spaces have been previously described to have mosquitoes [27] and hence represent some degree of risk for transmission when people are nearby. Unfortunately, we do not have data about the time people spend in those locations. We think that the number of infections happening while people are moving through these locations is likely to be negligible given the short duration of time people spend there, but this merits further work.
The distance that people travel to NH should not, however, be neglected when understanding urban spatial dynamics of transmission. The results of this model indicate that distance traveled is important to determine the level of clustering of cases, which is a measure of the level of spatial dependence of cases and hence of how cases are unevenly distributed in space (Fig 5) [24]. The clustering of cases considering the major role in transmission of NH is something that has not been explored before and deserves further exploration to understand its implication for disease control program design.
In this model we also included estimates of mosquito movement, where we found that high levels of mosquito migration induce high mosquito mortality, thereby reducing mosquito abundance and transmission (Fig 6). These results suggest that mosquito migration is carried out at a high cost for the individual and population, which supports the empirical observation that mosquitoes tend to stay in or close to the same location where they are breeding. Though our model has different parameterization and purpose from that previously reported by Reiner and colleagues, they also included both movement of people and mosquitoes and found that the former is the real force shaping dengue transmission as opposed to movement of mosquitoes, which diffuse it. Interestingly, they found that their model output matches real observations when mosquito movement is decreased and, under certain scenarios, when mosquito movement equals to zero [28].
In our model, a migrant mosquito travels to another location, irrespective of the suitability of that location (i.e., presence of water containers or established mosquito population), with a distance-based probability. If the new location is suitable for breeding, a new mosquito subpopulation is established, otherwise the migrant mosquito will die before reproducing. A migrant mosquito may be more attracted to travel to a location with water containers, as suggested by other studies [29], which is not included in the model. However, it is not known how likely it is for a migrating mosquito to travel longer distances in search of containers during migration if the closest location does not have any. This type of dynamics has never been studied before. Another variable not considered is the number of buildings, roads, and other urban features that might impose some limits to mosquito dispersal like those described previously, where all Ae. aegypti individuals were recaptured in the same block where they were released, unable to cross roads [30].
Though there are several mark-release-recapture studies recording long travel distances by Ae. aegypti [23], most of these studies are artificially releasing mosquitoes where they are forced to travel looking for a place to settle. While these studies are useful to evaluate the capacity of mosquitoes to fly in these conditions, they are not indicative of the propensity for already settled mosquitoes to travel the distances recorded in these studies.
According to results of this model, intra-urban dispersion of mosquitoes is not explained by mosquito migration. A plausible explanation is that dispersion is actively driven by human mobility. This would underscore the importance of human movement by suggesting that it is mainly responsible for the burden and dispersal of dengue through communities. Though the role of human mobility on intra-urban mosquito dispersion is hard to measure, there are several studies explaining the importance of human movement in dengue transmission at different space levels [11, 12, 28, 31–34].
Considering previous, novel results about the importance of NH environments in dengue transmission, we aimed to describe the implications of them across different urban spatial distributions and levels of human and mosquito movement, which has never been studied before. According to our model, those municipalities with NH locations clustered are slightly more protected from dengue spreading than those municipalities with NH randomly distributed in space. However, the main variable influencing dengue burden is the number of people moving between HH and NH. Together, these results reflect the importance of NH and human mobility between NH and HH spaces in dengue epidemiology. This underscores the importance of vector control in NH spaces, which is not currently implemented in most places. Finally, though people’s travel distance did not have a large impact on the number of cases, it is important for shaping spatial patterns, which can have implications for control activities and for local herd immunity.
Data Availability
All data produced are available online at github.com/vhpenagarcia/ABM_dengue
Acknowledgements
We would like to acknowledge Dr. Jason R. Andrews for the valuable input in this work. We would also acknowledge the fieldwork teams (in Kisumu: Joel Mbakaya, Samwel Ndire and Charles Adipo; in Ukunda: Said Lipi Malumbo, Paul S. Mutuku, Charles M. Ng’ang’a) who collected the data which this work was possible with.
References
