Household Characteristics, Residents’ Behaviors, and Seasonal Factors Associated With the Presence and Persistence of Aedes aegypti in a Dengue-Endemic Area Authors: Alejandro Cortés-Meda, Guadalupe Ponciano Rodríguez, Guadalupe Silvia García-de la Torre, Julián Everardo García-Rejón, Jorge Fernando Méndez-Galván Published: 2026 ABSTRACT Objective: This study aimed to evaluate the association between household characteristics, residents' preventive behaviors, seasonal factors, and the presence and persistence of Aedes aegypti in a dengue-endemic area. The findings are intended to identify modifiable household and behavioral factors that can inform targeted, seasonally adapted vector control strategies.Methodology: This prospective cohort study was conducted in San Pedro Pochutla, Oaxaca, Mexico. A total of 68 inhabited dwellings were selected and characterized on a single occasion according to their structure, degree of marginalization, vulnerability, and environment. Over a period of 12 months, seven entomological evaluations were performed indoors and outdoors to assess the presence, persistence, and seasonal patterns of Aedes aegypti.Results: A total of 476 visits were conducted, identifying 494 breeding sites positive for Aedes aegypti larvae or pupae. A total of 48.5% of the homes studied were positive on at least two occasions, with marked temporal variation, particularly during the rainy season (67.7%). Most breeding sites were located outdoors (96%), predominantly in controllable containers (73%). The most important habitats were drums, barrels/sinks/tubs/pools, cans/buckets, tires, and kitchen/washing utensils (72.8%). The characteristics of the dwelling that represented a risk for the presence and persistence of Aedes aegypti were the lack of mosquito nets on doors (OR=1.31, 95% CI=1.08-5.20), lack of pavilions on beds (OR=1.95, 95% CI=1.16-8.35), animals in the household (OR=2.27, 95% CI=1.19-7.63), and not applying insecticide indoors (OR=1.64, 95% CI=1.10-4.21).Conclusion: Household conditions, human behavior, and seasonal variation significantly influence the presence and persistence of Aedes aegypti. These findings can inform targeted vector control strategies, particularly focusing on modifiable risk factors and seasonal interventions. Introduction Aedes aegypti (Ae. aegypti) is considered the most important vector mosquito worldwide due to its role in transmitting multiple emerging arboviruses to humans, including dengue, Zika, chikungunya, yellow fever, and Mayaro virus. Among these, dengue is the most widespread globally [1-5]. This mosquito is found mainly in urban and suburban domestic environments, particularly in disadvantaged areas, where it finds optimal conditions for the development of its life cycle [6-9]. In Mexico, Ae. aegypti represents a major public health problem. It is estimated that nearly 60% of the national territory, where more than 50 million people live and where most of the agricultural, livestock, industrial, fishing, oil, and tourism centers of importance in our country are located, has conditions that favor the survival and distribution of this vector and the diseases it is capable of transmitting, with dengue the most important [10]. In 2023, Mexico reported the highest number of confirmed dengue cases to date, with 54,406 cases; of these, 25,570 (47%) corresponded to dengue with warning signs and severe dengue, with a case fatality rate of 0.79 per 100 cases [11]. The presence of Ae. aegypti has shown seasonal variations related to local ecology and meteorological parameters such as temperature, humidity, and rainfall [6,12]. Breeding sites of Ae. aegypti are predominantly found in domestic environments and are typically associated with artificial water-holding containers. These can be classified as follows: permanent or controllable (e.g., barrels, drums, cisterns, tanks, or buckets), which can be managed through larvicides and physical interventions; removable or disposable (e.g., tires, bottles, cans, and other containers); and natural (e.g., rock holes, coconut shells, tree hollows, and leaf axils). Any container capable of retaining water in areas where the mosquito is present may serve as a breeding site [2,10,13-24]. The life cycle of this mosquito is strongly influenced by household-level factors, such as inadequate water supply, lack of drainage systems, improper waste disposal, and poor housing conditions. In addition, a decreased perception of risk among the population negatively affects participation in vector control activities, such as the elimination of breeding sites, which is further aggravated by limitations in the financial resources allocated to vector control programs [25]. Currently, there is no widely effective antiviral treatment for dengue, and although vaccines exist, their use remains limited in many endemic settings. Therefore, prevention strategies rely primarily on reducing contact between humans and the vector [18,26]. Several studies have shown that the transmission of arboviruses is closely related to vector density, survival, and behavior [27-31]. Although socioeconomic status has been associated with the abundance of Ae. aegypti mosquitoes and dengue transmission, the probable driving factors that mediate this association, such as ecological factors specific to the household and understanding the spatiotemporal scale in which the encounter between mosquitoes and people occurs, have not been fully identified yet [32]. Materials and methods Methodology This is a prospective cohort study carried out in inhabited households in a neighborhood of San Pedro Pochutla, Pochutla district, La Costa region, Oaxaca, between September 2020 and August 2021 (Figure 1) [33]. The sampling strategy was based on the recommendations of the entomological guide for studies in the larval and pupal phase developed by the National Center for Preventive Programs and Disease Control and previous studies, which recommend selecting 10-25% of the available households to achieve adequate entomological precision [32,34]. Of the 204 households in the study area, 86 (42.2%) were randomly selected. A total of 68 households (33.3% of all households and 79.1% of those initially selected) completed the seven scheduled entomological evaluations and were included in the final analysis. The households were chosen through a simple random selection process. If the father or mother of the family was not found or refused to participate, the next house located on the right was chosen. Place of the study. This image was created by the authors of this study using Microsoft PowerPoint (Redmond, WA: Microsoft Corp.) and Paint 3D (Redmond, WA: Microsoft Corp.). Prior to the first entomological evaluation, and after obtaining written informed consent, a knowledge, attitudes, and practices (KAP) survey regarding dengue and Ae. aegypti was administered to the head of household of each participating dwelling. This baseline assessment aimed to characterize household-level behavioral factors potentially associated with vector presence. Study site San Pedro Pochutla is located between the 16°47’ north latitude and 96°28’ west longitude coordinates, at an altitude of 150 m above sea level. It is located approximately 230 km from the state capital. The climate is warm, subhumid with summer rains, average annual temperatures of 22-38°C, average humidity of 72% (64-80%), and average annual rainfall of 101 mm [35]. The population density of the municipality is 48.34 inhabitants/km2 . The housing conditions in the municipality are relevant because they show the need for basic services: nine (13%) do not have a connection to the public network for piped drinking water distribution, 42 (62%) depend mainly on the use of septic tanks, and only 26 (38%) are connected to the public network for waste disposal [35-37]. The study region is characterized by being hot and humid most of the year, with the dry period from January to May, November and December, and the rainy period from June to October (August and September being the months with the highest rainfall) [35]. This area is endemic to dengue, with an endemic profile that is usually seasonal, with large outbreaks from June to September, although cases can be observed at any time of the year. Circulation of serotypes one and two (dengue virus serotype 1 {DENV-1} and dengue virus serotype 2 {DENV-2}) has been observed, with a great predominance of the latter [38]. Entomological evaluations Prior to the first entomological assessment, a mapping of the study area (Chapingo Neighborhood) was carried out (Figure 1). A total of seven entomological assessments were carried out inside and outside the homes under study as follows: three during the rainy season (September and October 2020 and August 2021) and four in the dry season (November 2020 and January, February, and March 2021). For each of these months, the Stegomyia indices were calculated, which are essential for monitoring the Ae. aegypti population [6,39-41]. Those calculated in this project are the following: Positive House Index (PHI) = (households with positive containers/total households explored) × 100; Positive Container Index (PCI) = (positive containers/total containers with water explored) × 100; Breteau Index (BI) = (positive containers/explored houses) × 100. The operational criteria for interpretation of PHI, PCI, and BI of the Pan American Health Organization (PAHO) and the methodological guide for entomological studies in the larval and pupal phase of the National Center for Preventive Programs and Disease Control (CENAPRECE) and other publications were followed (Table 1) [6,42]. Table 1Interpretation of entomological indices.Own elaboration based on the methodological guide for entomological studies in the larval and pupal stages [34].Operational control levelPositive Household IndexPositive Container IndexBreteau IndexOptimal<1%<0.5%1-4.99Good1-2.99%0.5-1.99%5-9.9Alarm3-4.99%2-4.99%10-14.9Emergency≤5%≤5%≤15 Interpretation of entomological indices. Own elaboration based on the methodological guide for entomological studies in the larval and pupal stages [34]. The containers inside and outside each household were inspected for mosquito larvae and pupae and classified (controllable and disposable). All larvae (in the third and fourth stage) and pupae from the positive containers were collected using pipettes and scoops, counted, and recorded on forms unique to each household [43]. They were identified according to standard taxonomic keys [44]. Classification of breeding sites (breeding grounds) They were classified as follows: controllable breeding grounds - these are containers that are present throughout the year, in use, and that due to their characteristics cannot be moved from the home. Disposable breeding grounds - all those containers that are not used, that are considered as garbage, and should be eliminated from the household, since, due to carelessness or the effect of rain, they retain water [40]. Data analysis The data were analyzed using the statistical package Stata version 15 (College Station, TX: StataCorp LLC). The studied households and the behavior of their residents were characterized; the proportion of breeding sites positive for larvae and pupae was determined, as well as their location (inside or outside the homes), type (controllable or removable), and seasonal behavior. To improve the analysis, the households were classified according to their persistence for larvae and pupae as follows: persistent - households positive for Ae. aegypti breeding sites (larvae and pupae), for two or more months; non-persistent - households positive for Ae. aegypti breeding sites (larvae and pupae), for one month or never positive. Persistence was operationally defined as household positivity in two or more visits. Although there is no universally accepted cutoff for defining household persistence of Aedes aegypti, this criterion was established a priori based on epidemiological considerations and the study's longitudinal design. Repeated detection of immature Aedes aegypti stages during independent household visits was considered indicative of recurrent household infestation rather than an isolated observation. This definition is consistent with previous longitudinal studies that conceptualize household persistence as the repeated presence of the mosquito over time rather than a single positive inspection [2]. Using the X2 test and Fisher's exact test, the association between the characteristics of the homes, KAP of the residents, and entomological indices with persistence of the mosquito Ae. aegypti was established. Finally, a multivariable logistic regression analysis was performed to evaluate the association between household characteristics and the presence and persistence of Aedes aegypti. The multivariable logistic regression model was constructed using a forward selection procedure. Initially, variables showing a statistically significant association in the bivariate analysis (p<0.05) were considered. However, variable selection was not based solely on statistical significance. During model construction, the epidemiological relevance of the variables and the available evidence from the scientific literature were also considered to evaluate their potential role as confounding factors, even if they did not reach statistical significance in the bivariate analysis. Before fitting the model, potential multicollinearity among the independent variables was assessed using the variance inflation factor (VIF). All VIF values were below 5, indicating no evidence of relevant multicollinearity and supporting the simultaneous inclusion of the selected variables in the multivariable model. Model diagnostics also included an assessment of goodness of fit using the Hosmer-Lemeshow test. Finally, the most parsimonious model was selected, retaining only those variables that remained statistically significant and were epidemiologically consistent. Results Seasonal and climatic factors at the study site During the studied period, an average annual temperature of 27.3°C, rainfall of 718 mm, and humidity of 64.2% were recorded (Figure 2). Average temperature, humidity, and precipitation per epidemiological week. *Rainy season. **Dry season. This image was created by the authors of this study using Microsoft PowerPoint (Redmond, WA: Microsoft Corp.) and Paint 3D (Redmond, WA: Microsoft Corp.). Temporal trends by epidemiological week, follow-up month, and season (rainy and dry) in average temperature (maximum and minimum), rainfall, humidity, and the total number of water containers harboring Ae. aegypti larvae and pupae in the Chapingo neighborhood of San Pedro Pochutla, Oaxaca. Entomological indicators As a result of the entomological evaluations, the House Index (HI), Container Index (CI), and Breteau Index (BI) were calculated to assess the levels of Ae. aegypti infestation in the study area (Figure 3, Table 2). Throughout the study period, the average values of all three indicators remained above the levels generally considered acceptable for vector control. The highest HI value was recorded in September 2020 (70.6%), whereas the highest CI and BI values were observed in August 2021 (46.8% and 280.9, respectively). Monthly variation in the House Index (HI), Container Index (CI), and Breteau Index (BI) during the study follow-up period. *Rainy season. **Dry season. One-sample T-test. This image was created by the authors of this study using Microsoft PowerPoint (Redmond, WA: Microsoft Corp.) and Paint 3D (Redmond, WA: Microsoft Corp.). Table 2Monthly values of the House Index (HI), Container Index (CI), and Breteau Index (BI) during the study period.*Rainy season. **Dry season.IndexSeptember-20*October-20*November-20**June-21**February-21**March-21**August-21*AverageHouse Index (HI)70.6%41%30.2%25.0%33.8%20.6%47.1%38.4%Container Index (CI)12.9%8.7%15.2%6.9%10.6%3.3%46.8%13.4%Breteau Index (BI)117.686.889.745.67530.9280.9103.8 Monthly values of the House Index (HI), Container Index (CI), and Breteau Index (BI) during the study period. *Rainy season. **Dry season. The overall average values obtained during the study period were 38.4% for the HI, 13.4% for the CI, and 103.8 for the BI. In general, HI and BI values were higher during the rainy season months (September and October 2020, and August 2021) than during the dry season months (November 2020, and February, March, and June 2021). Notably, August 2021 showed the highest levels of entomological infestation, with the maximum values recorded for both the CI and BI. These findings indicate a greater proportion of positive containers and a higher level of household infestation during this period, demonstrating a high presence and persistence of Ae. aegypti in the study area. Classification of positive containers During the study, 14,484 breeding sites were found, of which 11,010 (76%) were negative since they did not contain water, 2,980 (20.6%) were negative with water, and 494 (3.4%) were positive for Ae. aegypti larvae and pupae. Of the 494 positive breeding sites, 474 (96%) were found in the peridomicile. Of these, 346 (73%) were controllable and 128 (27%) disposable. When analyzing peridomicile breeding sites by season, 321 (67.7%) were found during the rainy season, of which 205 (63.8%) were controllable. Among these, barrels, kitchen and washing utensils, dishes, tanks/sinks/tubs/basins, and buckets or containers were the most frequent, representing 67.8% (139) of the controllable breeding sites. While the remaining 153 (32.3%) were identified in the dry season; of these, 141 (92.2%) were controllable, with barrels and tanks/sinks/tubs/basins being the most common, with 260 (54%). Of the disposable breeding sites, tires represented 57.8% (74) (Table 3). However, it is important to restate that, beyond the amount of rainfall, humidity had a more significant impact on the productivity of positive breeding sites. Table 3Positive hatcheries for Aedes aegypti larvae and pupae.Descriptive statistical analysis was performed using absolute frequencies “n” and percentages “(%).” No inferential statistical tests were applied to this table.C: controllable; D: disposableDateSeptember 2020October 2020November 2020June 2021February 2021March 2021August 2021Total, n (%)Type of hatcherynnnnnnnCDCDCDCDCDCDCDExterior of the householdTank/deposits/basins/water stack7-3-9-7-9-9-18-62 (13)Barrels21-7-9-6-20-7-27-97 (20.5)Water tanks/cisterns/wells4-4-2-5-7-3-3-28 (6.0)Cans/buckets612-324-412-132058 (12.2)Tires213-30111-3---73088 (18.6)Water dishes4-1-7---1-----13 (2.7)Kitchen and washing utensils111-2-------33240 (8.4)Useless things (stuff)-51-514-----2-18 (3.8)Diverse small <5 L-8-1-7--1---5325 (5.3)Bottles/tins-11-3-------16-21 (4.4)Others4-512-4-1---7-24 (5.0)Exterior total49292532431131046121013155474 (100) Positive hatcheries for Aedes aegypti larvae and pupae. Descriptive statistical analysis was performed using absolute frequencies “n” and percentages “(%).” No inferential statistical tests were applied to this table. C: controllable; D: disposable For the following analyses, it was decided to exclude natural and indoor breeding sites, since these together only represent 4% (20) of the total. Therefore, the total number of breeding sites with which the persistence of Ae. aegypti was measured decreased from 494 to 474. Persistence In total, 68 households accomplished a complete follow-up, of which 33 (48.5%) were considered persistent, since they tested positive two or more times throughout the study (Figure 4). Spatial distribution of households with persistent and non-persistent Aedes aegypti positivity in the Chapingo neighborhood during the study period. This image was created by the authors of this study using Microsoft PowerPoint (Redmond, WA: Microsoft Corp.) and Paint 3D (Redmond, WA: Microsoft Corp.). Demographic characteristics The sociodemographic characteristics evaluated included sex, age, number of inhabitants per household, number of bedrooms per household, employment status, occupation, number of working hours per day, monthly income, housing tenure, social security coverage, distance between the household and the nearest health facility, and religion. Among the sociodemographic characteristics analyzed, the number of inhabitants per household, employment status, total family income, and social security coverage showed a significant association with the persistence of Ae. aegypti (Table 4). Table 4Association between demographic characteristics and household persistence of Aedes aegypti.Descriptive statistical analysis was performed using absolute frequencies (n) and percentages (%). Inferential statistical analysis was conducted using the chi-square (χ²) test or Fisher's exact test when expected cell frequencies were less than five. A p<0.05 was considered statistically significant.MXN: Mexican pesoFactorVariablesNon-persistent, n (%)Persistent, n (%)Total, n (%)X2 Fisher’s exactp-ValueInhabitants1-316 (57.1)12 (42.9)28 (100)-6.710.0014-615 (46.9)17 (53.1)32 (100)7 or more4 (50.0)4 (50.0)8 (100)Employment situationEmployee17 (56.7)13 (43.3)30 (100)6.59-0.01Unemployed16 (44.4)20 (55.6)36 (100)Family’s total income<5000 MXN21 (41.2)30 (58.8)51 (100)-7.920.005≥5000 MXN12 (75.0)4 (25.0)16 (100)Social SecurityYes17 (65.4)9 (34.6)26 (100)7.76-0.005No17 (41.5)24 (58.5)41 (100) Association between demographic characteristics and household persistence of Aedes aegypti. Descriptive statistical analysis was performed using absolute frequencies (n) and percentages (%). Inferential statistical analysis was conducted using the chi-square (χ²) test or Fisher's exact test when expected cell frequencies were less than five. A p<0.05 was considered statistically significant. MXN: Mexican peso Households with four to six inhabitants showed a higher proportion of persistence, with 17 (53.1%) households classified as persistent, compared with 12 (42.9%) households among those with one to three inhabitants. Likewise, households in which the head of household was unemployed showed a higher proportion of persistence, with 20 (55.6%) persistent households, compared with 13 (43.3%) among households in which the head of household was employed. Similarly, persistence was more frequent among households without social security coverage, with 24 (58.5%) persistent households, compared with nine (34.6%) among households with social security coverage. In addition, households with a monthly family income below 5,000 Mexican pesos showed a higher proportion of persistence, with 30 (58.8%) persistent households, whereas households with a monthly income of 5,000 Mexican pesos or more showed only four (25.0%) persistent households (Table 4). Housing and resident factors associated with the presence and persistence of Ae. aegypti Various housing and resident characteristics were evaluated, including wall structure, vegetation surrounding the household, percentage of shade in the patio, ventilation, construction materials, kitchen characteristics, cooking fuel, the number of doors and windows leading outdoors, basic services, preventive practices, and animal ownership. Variables significantly associated with the persistence of Ae. aegypti included the absence of mosquito screens on doors leading outdoors (Fisher’s exact test=2.38; p=0.02), absence of bed canopies or bed nets (χ²=2.59; p=0.04), participation in vector control campaigns (χ²=1.35; p=0.03), presence of animals inside the household (χ²=1.45; p=0.003), type of animal (χ²=2.06; p=0.02), absence of a corral or designated animal enclosure (χ²=3.64; p=0.02), number of exterior doors (χ²=1.99; p=0.04), and use of insecticide inside and outside the household (Fisher’s exact test=1.73; p=0.044 and Fisher’s exact test=1.71; p=0.025, respectively). Households without mosquito screens on exterior doors showed 32 (54.2%) persistent households, compared with two (25.0%) households with mosquito screens on exterior doors. Likewise, persistence was observed in 23 (59.0%) households without bed canopies or bed nets, compared with 10 (35.7%) households that reported having them. Persistence was also observed in 28 (50.9%) households with animals inside the household, compared with five (41.7%) households without animals. Among households with animals, persistence was more frequent in those with domestic animals, with 21 (56.8%) persistent households, compared with seven (38.9%) households with farm animals. Similarly, households without a corral or designated animal enclosure presented 20 (64.5%) persistent households, whereas eight (33.3%) persistent households were observed among households with a corral or enclosure. Participation in vector control campaigns was associated with persistence, with 14 (53.8%) persistent households among those reporting participation, compared with 19 (46.3%) persistent households among those that did not participate. Regarding the number of exterior doors, persistence was observed in eight (38.1%) households with one exterior door, 11 (55.0%) households with two exterior doors, and 13 (59.1%) households with three or more exterior doors. The use of insecticide inside and outside the household was also significantly associated with the persistence of Ae. aegypti. Persistence was observed in 30 (60.0%) households that did not apply insecticide indoors, compared with four (23.5%) households that reported indoor insecticide use. Similarly, persistence was observed in 30 (60.0%) households that did not apply insecticide outdoors, compared with four (23.5%) households that reported outdoor insecticide application. On the other hand, drainage services, history of dengue within the household, presence of mosquito screens on windows, location of animal drinking containers, wall structure, flooring material, wall material, roof material, and number of windows to the exterior were not significantly associated with the persistence of Ae. aegypti (Table 5). Table 5Association between household characteristics and household persistence of Aedes aegypti.Descriptive statistical analysis was performed using absolute frequencies “n” and percentages “%.” Inferential statistical analysis was conducted using the chi-square (χ²) test and Fisher’s exact test. A p<0.05 was considered statistically significant.FactorVariablesNon-persistent, n (%)Persistent, n (%)Total, n (%)X2 Fisher’s exactp-ValueDrainagePublic sewer6 (33.3)12 (66.6)18 (100)2-0.08Septic tank28 (57.1)21 (42.9)49 (100)History of dengue in the houseNo23 (46.9)26 (53.1)49 (100)1.78-0.36Yes11 (61.1)7 (38.9)18 (100)Mesh or mosquito screens on windowsNo27 (54.0)23 (46.0)50 (100)1.43-0.08Yes7 (41.2)10 (58.8)17 (100)Mesh or mosquito screens on doorsNo27 (45.8)32 (54.2)59 (100)-2.380.02Yes6 (75.0)2 (25.0)8 (100)Pavilions on bedsNo16 (41.0)23 (59.0)39 (100)2.59-0.04Yes18 (64.3)10 (35.7)28 (100)Participation in vector control campaignsNo22 (53.7)19 (46.3)41 (100)1.35-0.03Yes12 (46.2)14 (53.8)26 (100)Animals in the houseNo7 (58.3)5 (41.7)12 (100)1.45-0.003Yes27 (49.1)28 (50.9)55 (100)Type of animalDomestic16 (43.2)21 (56.3)37 (100)2.06-0.02Farm11 (61.1)7 (38.9)18 (100)Corral for animalsNo11 (35.5)20 (64.5)31 (100)3.64-0.02Yes16 (66.7)8 (33.3)24 (100)Location of animal drinking containersInside2 (22.2)7 (77.8)9 (100)-3.50.13Out14 (50.0)14 (50.0)28 (100)Wall structuresContinuous20 (52.6)18 (47.4)38 (100)1.19-0.72Discontinuous14 (48.3)15 (51.7)29 (100)Flooring materialCement29 (56.9)22 (43.1)51 (100)2.9-0.07Other materials5 (31.2)11 (68.8)16 (100)Wall materialBricks/blocks and other material5 (38.5)8 (61.5)13 (100)0.54-0.32Discarded material29 (53.7)25 (46.3)54 (100)Roof materialConcrete16 (50.0)16 (50.0)32 (100)-1.50.68Concrete/metal sheet4 (40.0)6 (60.0)10 (100)Others14 (56.0)11 (44.0)25 (100)Exterior doors113 (61.9)8 (38.1)21 (100)1.99-0.0429 (45.0)11 (55.0)20 (100)Three or more9 (40.9)13 (59.1)22 (100)Windows to the exterior1-213 (48.2)14 (51.8)27 (100)1.02-0.133-49 (52.9)8 (47.1)17 (100)Five or more10 (47.6)11 (52.4)21 (100)Use of insecticide inside the houseNo20 (40.0)30 (60.0)50 (100)-1.730.044Yes13 (76.5)4 (23.5)17 (100)Use of insecticide to the outside of the houseNo20 (40.0)30 (60.0)50 (100)-1.710.025Yes13 (76.5)4 (23.5)17 (100) Association between household characteristics and household persistence of Aedes aegypti. Descriptive statistical analysis was performed using absolute frequencies “n” and percentages “%.” Inferential statistical analysis was conducted using the chi-square (χ²) test and Fisher’s exact test. A p<0.05 was considered statistically significant. Characteristics of the home and its inhabitants for the persistence of Ae. aegypti Table 6 presents the results of the multivariable logistic regression analysis. The factors independently associated with the persistence of Aedes aegypti were the absence of mosquito screens on doors leading outdoors (OR=1.31; 95% CI: 1.08-5.20), the absence of bed nets or bed canopies (OR=1.95; 95% CI: 1.16-8.35), the presence of animals within the household (OR=4.27; 95% CI: 1.19-9.63), the presence of domestic animals (OR=2.01; 95% CI: 1.15-11.36), the absence of a corral or designated enclosure for animals (OR=2.46; 95% CI: 1.14-11.19), having two doors leading outdoors (OR=2.30; 95% CI: 1.22-10.71), and not applying insecticide inside the household (OR=1.64; 95% CI: 1.10-4.21). Table 6Multivariable logistic regression analysis of factors associated with household persistence of Aedes aegypti.Multivariable logistic regression analysis (n=68). All variance inflation factor (VIF) values <5. The Hosmer-Lemeshow test indicated an adequate model fit p=0.28.z: Wald z statistic; ref: reference categories were used for comparison.FactorVariablesOR95% CIZ-valuep-ValueDrainagePublic sewer (ref.)--0.448Septic tank0.470.10-2.16-0.96Mesh or mosquito screens on windowsNo (ref.)0.460.06-6.56-0.650.601Yes--Mesh or mosquito screens on doorsNo1.311.08-5.201.960.036Yes (ref.)--Pavilions on bedsNo1.951.16-8.352.540.011Yes (ref.)--Participation in campaignsNo (ref.)--0.167Yes1.360.88-2.101.38Animals in the householdNo (ref.)--0.006Yes2.271.19-7.632.72Type of animalDomestic2.011.15-11.361.990.047Farm (ref.)--Corral for animalsNo2.461.14-11.192.030.042Yes (ref.)--Flooring materialCement (ref.)--0.18Other material2.880.61-13.621.34Exterior doors1 (ref.)---22.31.22-10.711.980.0483 or more1.520.27-8.570.470.26Use of insecticide inside the houseNo1.641.10-4.212.310.021Yes (ref.)--Use of insecticide to the outside of the houseNo1.390.95-2.891.160.25Yes (ref.)-- Multivariable logistic regression analysis of factors associated with household persistence of Aedes aegypti. Multivariable logistic regression analysis (n=68). All variance inflation factor (VIF) values <5. The Hosmer-Lemeshow test indicated an adequate model fit p=0.28. z: Wald z statistic; ref: reference categories were used for comparison. These findings suggest that specific structural characteristics of the household, animal management practices, and indoor vector control measures play an important role in the persistence of Aedes aegypti in households in the Chapingo neighborhood of San Pedro Pochutla, Oaxaca. The model was adjusted for household size, monthly household income, and social security status; however, none of these variables showed a statistically significant association with the outcome. Discussion Ae. aegypti is widely recognized as a domestic vector with a strong preference for human blood feeding, even in the presence of alternative hosts [1]. Its high competence for transmitting viral infections, particularly dengue, underscores its global public health importance. This study identified key social, environmental, and behavioral factors associated with the presence and persistence of Ae. aegypti. The entomological methodology, based on the identification of positive breeding sites and the collection of larvae and pupae, proved effective for documenting vector presence, consistent with previous studies [23,45]. Environmental conditions, particularly seasonality, played a critical role in vector dynamics. A higher proportion of positive breeding sites was observed during the rainy season, with a total of 322 (68%), which is consistent with findings reported in previous studies [46,47]. High relative humidity, along with constant temperature ranges (23°-31°C), likely contributed to increased vector proliferation by accelerating development rates, shortening reproductive cycles, and enhancing survival [20,46-52]. Consistent with WHO recommendations, routine entomological surveillance remains essential to monitor fluctuations in vector density and geographic distribution, thereby informing risk assessment and control strategies [6]. In this context, the entomological indices used (PHI, PCI, and BI) demonstrated sensitivity to seasonal variations, with significantly higher values during the rainy season. Although these indices have been questioned due to their historical origin and the absence of well-defined dengue transmission thresholds [42,53,54], they remain useful for identifying priority intervention targets and understanding vector ecology [22,55]. The predominance of immature stages in the peridomiciliary environment aligns with previous reports [2,40]. Only 11 (2.4%) breeding sites were located indoors, likely due to frequent domestic use and cleaning of water containers, which disrupts larval development. Additionally, 48 (70%) households reported covering indoor water storage containers, further reducing indoor breeding potential [2,56]. The high reproductive activity in the home environment tells us about the adaptation of Ae. aegypti to indoor and outdoor habitats, which allows it to increase its reproduction opportunities and increase the rate of larval development [1,57]. This may have important implications for the transmission of dengue and other arboviruses. According to other studies, most of the positive breeding sites are disposable, which was not found in this study, since of the total number of positive breeding sites identified, 346 (73%) were controllable [40,47,58]. The importance of controllable containers was maintained in both seasons; 303 (64%) of the total identified in the rainy season were of this type, while in the dry season it was 436 (92%). This requires that vector mitigation actions place greater interest in the habitats of larvae in the home environment [2]. The main positive breeding sites outside the houses during the rainy season were tires, barrels, buckets, and kitchen or washing implements. In the present work, the most relevant breeding sites were tanks/sinks, which represented 67% (318) of the total. Meanwhile, in the dry season, the most relevant breeding sites were barrels and tanks/tubs/sinks, which together represented 50%. Results that are comparable with other studies [14,18,23,59,60]. The high productivity of permanent containers may be due to the need to store water by the population for carrying out their daily activities, since the supply is irregular. In 65 (95%) households studied, water storage in large outdoor containers was reported. Its importance is due to the capacity to contain sufficiently large volumes of water for considerably longer periods than are suitable for complete larval development; therefore, they can be considered as key containers in both seasons [2,58]. Tires were the most important disposable breeding site throughout the study; their importance as a key breeder has been widely reported [2,16,47,58]. The high proportion of persistent households 33 (48.5%) may be explained by underlying socioeconomic and structural conditions, including inadequate water supply, poor sanitation, and suboptimal housing conditions. These factors have been consistently associated with increased vector proliferation [2,6-9]. Preventive measures were limited in the study population. A high percentage of households lacked mosquito nets on doors (59, 86%) and windows (51, 75%), and more than half did not use bed nets. Furthermore, 51 (75%) reported not using insecticides or repellents, and 41 (60%) had never participated in vector control campaigns. These findings suggest low risk perception and limited adoption of preventive behaviors, which may contribute to sustained transmission risk. The presence of domestic animals inside households may also facilitate vector persistence by providing additional blood sources and creating breeding opportunities through water containers such as troughs [2,15]. This study has limitations that should be considered when interpreting the findings. First, the relatively small sample size may have reduced the precision and stability of the estimates obtained from the multivariable logistic regression model. In addition, although the number of households lost to follow-up was limited, their exclusion may have affected the representativeness of the analytical sample and introduced potential selection bias. As with any observational study, the possibility of residual confounding cannot be completely excluded, despite adjustment for potential confounding variables in the multivariable analysis. Second, this study was conducted in a single dengue-endemic community, which may limit the generalizability of the findings to areas with different ecological, climatic, or socioeconomic characteristics. Third, household persistence was operationally defined as positivity in two or more visits. Although this definition was established a priori based on epidemiological considerations and the longitudinal design of the study, no universally accepted definition of household persistence currently exists, and alternative classification criteria may be used in future studies. Finally, repeated household visits may have influenced participants' behaviors over time, and temporary interruptions in data collection during the COVID-19 pandemic may have affected seasonal comparisons. Accordingly, the observed associations should be interpreted with caution and confirmed in prospective studies with larger sample sizes conducted in different epidemiological settings. Despite these limitations, this study has several strengths. The prospective cohort design, repeated household evaluations over 12 months, and the integration of entomological, household, behavioral, and seasonal data enabled a comprehensive assessment of factors associated with the presence and persistence of Aedes aegypti. These findings may help inform targeted vector control strategies and provide a basis for future studies with larger sample sizes and broader geographic representation. Conclusions This study from Mexico comprehensively characterizes the habitats of Ae. aegypti and shows that controllable household breeding sites are the main drivers of vector persistence in both rainy and dry seasons. In tropical endemic settings, relative humidity appears to be a key year-round environmental factor, even more stable than temperature and rainfall, favoring vector survival. Additionally, low household risk perception and free indoor access of domestic animals may increase domiciliary persistence. These findings support the development of a more focused entomological surveillance approach and provide evidence to guide context-specific prevention and vector control strategies that reduce human-mosquito contact in tropical endemic ecosystems.