Determinants of urban mosquito population density and community responses: A cross-sectional study

Version Changes

Revised. Amendments from Version 3

Thanks to the reviewer comments, this revised version incorporates few conceptual, methodological, and analytical enhancements compared to the previously published version. The Introduction has been expanded to include a dedicated paragraph on coastal urban vector ecology, highlighting the unique ecological and urbanization-driven determinants of mosquito proliferation in Mangalore, supported by recent evidence from other Indian coastal cities. A new conceptual framework section has been added using the eco-social framework to better situate the study within contemporary social-ecological theory. The research questions and objectives have been reorganized into clearly defined primary and secondary components for improved alignment with the study aims. Correspondingly, the Abstract has been mildly revised to reflect these refinements in the objective structure. The Methods section has been substantially strengthened by adding details on questionnaire validation procedures, including pilot testing, and Cronbach’s alpha for internal consistency. Operational definitions have been standardized and expanded for key variables such as mosquito presence, nuisance scores, self-reported mosquito-borne disease, and water stagnation. The sampling strategy has been clarified by describing purposive ward selection, random street initiation, random starting-point selection, and acknowledging the exclusion of apartment households. Analytical reporting has been strengthened by providing clearer description and presentation of the already existing multivariable logistic regression models, which used to identify independent predictors. The Results section has been revised to avoid causal language, with all tables reformatted to include clear titles, denominators, consistent n (%), 95% confidence intervals, crude and adjusted odds ratios, and corrected p-values (which were available in the previous version). The Discussion has been strengthened with expanded strengths and limitations—including seasonality, selection bias, and self-reporting—and enhanced public-health and policy implications. A few additional references have also been incorporated to support the expanded introductory content.

Abstract

Background

Vector-borne diseases transmitted by various arthropods account for approximately 17% of the global burden of infectious diseases. These arthropods, especially mosquitoes, are particularly rampant in Mangalore because of the humid coastal climate and scaling urbanization.

Objectives

To identify key environmental and household determinants of mosquito presence in urban Mangalore, to assess household-level prevention practices, and to evaluate community perceptions and self-reported disease burden towards mosquito-borne diseases.

Methods

The study involved households in selected wards of the urban field practice area of the Department of Community Medicine, a teaching and service field area under the Mangalore City Corporation, selected through convenience sampling. Data were collected using a semi-structured questionnaire covering sociodemographic details, mosquito proliferation, breeding determinants, behavioral measures, perception of mosquito control, and self-reported cases of mosquito-borne diseases. The data were analyzed using Jamovi version 2.6.26.

Results

Among 95 respondents (70.5% female and 94.8% literate), 42.1% reported an increase in mosquito breeding sites over the past year, 69.4% recognized the rainy season as the peak period of mosquito activity. Water stagnation [74.7% (95% CI: 64.8–83.1)] and ongoing construction activity [32.4% (95% CI: 21.8–44.1)] emerged as significant environmental determinants of higher mosquito density which was not statistically significant. A large majority of households (91.6%) reported using chemical measures for mosquito prevention, while 92.6% of participants were aware of mosquito-borne diseases. Despite this, nearly one-third (29.4%) of respondents had experienced a mosquito-borne illness in the preceding year, with 71.4% dengue infection. The use of mosquito repellents was paradoxically associated with a higher prevalence of mosquito-borne diseases (OR = 3.7; 95% CI: 1.4–9.6; p = 0.024).

Conclusion

Although awareness and preventive measure uptake were high, gaps remain in consistent environmental control and municipal interventions. Strengthening local authority action on water stagnation and construction-site management is essential for sustainable vector control.

Introduction

Vectors are defined as organisms that transmit infectious pathogens from humans to humans or from animals to humans. Common vector-borne diseases, including malaria, lymphatic filariasis, dengue, chikungunya, West Nile fever, yellow fever, Chagas disease, bubonic plague, and leishmaniasis, are transmitted by arthropod vectors. ¹ Together, these diseases account for approximately 17% of all infectious diseases worldwide and are responsible for nearly 700,000 deaths annually. ²

Mosquitoes are among the most prominent arthropod vectors, representing a significant portion of the vector-borne disease burden, with over 80% of the global population at risk. ¹ Mosquitoes are arthropods of medical importance under the class Insecta and are further divided into Anopheline and Culicine mosquitoes. Anopheline mosquitoes are the primary vectors of malaria; they generally exhibit nocturnal biting habits (typically bites between 10 PM and 4 AM) and indoor resting behavior and breed in clean, sunlit water sources. According to the World Health Organization (WHO), malaria alone accounted for approximately 249 million cases globally in 2023, with 94% occurring in the WHO African Region. ³ India contributes nearly 52% of malaria cases outside sub-Saharan Africa and represents approximately 79% of the malaria burden within the WHO Southeast Asia Region. ^{4, 5}

Culicine mosquitoes, notably Aedes aegypti and Aedes albopictus, are highly domesticated; they breed in water-filled containers in domestic and peridomestic areas and bite primarily during the day and early evening. These species transmit viral infections such as dengue, chikungunya, and Zika, which contribute to tens of thousands of deaths per year despite causing hundreds of millions of infections. ^{6– 8} Recent WHO global dengue surveillance, from January to November 2024, the total number of dengue cases was 13,860,025, with a total of 9990 deaths. ^{9, 10} The first chikungunya outbreak in India occurred in the 1960s, followed by a period of dormancy until a major resurgence in 2006, which affected 13 states in the country. ^{11, 12}

India’s vulnerability to mosquito-borne diseases is exacerbated by its eco-socio-demographic conditions, making it a major public health concern. Karnataka, a southern state in India, faces a significant burden of mosquito-borne diseases including malaria, dengue, lymphatic filariasis, and Japanese encephalitis, with the prevalence varying across districts. In 2010, Karnataka reported 1,09,118 malaria cases, 28,065 of which were attributed to Plasmodium falciparum. ¹³

The coastal cities of Mangalore and Udupi together account for approximately 72% of the malaria cases reported in Karnataka. ¹⁴ In particular, Mangalore—a coastal town in the Dakshina Kannada district with frequent heavy rainfall and high humidity—provides an ideal environment for mosquito proliferation. An entomological survey conducted in Mangalore taluk identified 26 mosquito species across six genera, with an annual parasite index of 10–12, indicating that the area was malaria endemic. ¹⁵

Coastal urban vector ecology, as seen in malaria-endemic cities such as Surat, is strongly influenced by relative humidity, which plays a critical role in the survival and population expansion of urban malaria vectors such as Anopheles stephensi. This urban mosquito thrives in densely built environments, utilizing artificial containers and construction-site water for breeding rather than depending on rainfall. Consequently, transmission intensity remains closely linked to relative humidity levels, with values above approximately 60% providing highly favorable conditions for sustained malaria transmission. ¹⁶ Mangalore, situated in a humid tropical coastal region, experiences year-round mosquito presence that peaks during the rainy season, with stagnant water and inadequate sanitation driving high vector densities. Rapid urbanization—characterized by extensive construction, poor drainage, and deteriorating road conditions—has further sustained the endemicity of mosquito-borne diseases in the area. ¹⁷

Numerous studies have identified environmental, socioeconomic, and behavioral determinants of mosquito proliferation and vector-borne disease transmission. Aquatic habitats, such as pools, streams, and water-filled containers, are critical breeding sites for mosquitoes. While many Anopheles species breed in relatively clean, fresh water, this is not universal. Notably, the urban vector Anopheles stephensi, which has become invasive in several regions including the Arabian Peninsula and the Horn of Africa, is well adapted for breeding in domestic water containers and polluted urban habitats. ¹⁸ It is also the predominant urban malaria vector in many parts of India, including coastal Karnataka. ¹⁹ Furthermore, behavioral adaptations among Anopheles species have been documented in response to vector control interventions, with some species shifting toward earlier or greater diurnal biting activity following indoor residual spraying (IRS) and the deployment of long-lasting insecticidal nets (LLINs). ²⁰

A study by Wilke et al. (2020) in Miami-Dade Florida identified a few of the common aquatic habits that are responsible for harboring 80% of all immature Ae. Aegypti and increasing their proliferation in the presence of those aquatic habitats. ²¹ Similarly, Prashanthi et al. (2007) reported that Anopheles breeds in pools and streams, where people living in close proximity are at high risk of malaria and its transmission. ²² These studies have also revealed that socioeconomic factors exacerbate the vulnerability to malaria, as economically marginalized populations often lack access to anti-mosquito measures, such as mosquito nets or repellents, and may follow age-old traditional practices, such as sleeping outdoors at night amid peak mosquito activity.

The use of preventive measures, such as effective lids over water storage containers and frequent emptying of containers, significantly reduces the incidence of arthropod proliferation, especially in Ae. aegypti. ²³ Studies have shown a linear relationship between growing populations, rising socioeconomic status, and increased mosquito proliferation, particularly in economically marginalized densely populated areas vulnerable to dengue. ²⁴ A study by Srividya et al. (2018), through logistic regression analyses, indicated that tiled and concrete dwellings increased the likelihood of an area becoming a dengue hotspot by 2.0 and 2.9 times, respectively, due to the conducive breeding environment. ²⁵

There is a significant relationship between rapid, unplanned urbanization and the proliferation of mosquitoes. Mangalore has experienced dramatic urbanization in recent years, and these unplanned disorganized cities aggravate mosquito proliferation, especially in Aedes aegypti by creating artificial breeding grounds, such as stagnant water pools, and increasing disease transmission. ²⁶ Climate change also has a considerable effect on vector proliferation. It reduces larval development time and rapidly increases mosquito populations. This also leads to a reduction in the extrinsic incubation period of pathogens in mosquitoes, thereby increasing their infectiousness. ²⁶

Community knowledge and behavior are critical for effective vector control. A study by Garbin et al. (2021) revealed that while 76% of respondents believed that their neighborhood was likely to be infected by a disease spread by mosquitoes, but no action was taken by them, highlighting a gap between awareness and actions. ²⁷ Another study by Madeira et al. (2002) demonstrated that didactic interventions among schoolchildren increased knowledge about mosquito breeding sites and vector proliferation, leading to heightened awareness. ²⁸

Various determinants of mosquito proliferation have been identified across different studies, and the present study builds on this evidence by exploring the primary objective of examining the specific environmental, socioeconomic, and behavioral factors driving mosquito proliferation in Mangalore and simultaneously fulfilling its secondary objectives of assessing community measures to mitigate vector-borne disease risk. This study also adopts the eco-social framework, which links environmental conditions, social structures, and community behaviours to disease risk. ²⁹ This framework clarifies how structural and ecological pathways sustain mosquito proliferation in Mangalore and influence community vulnerability to vector-borne diseases. By examining vector nuisance, disease prevalence, and community engagement, this study aligns with Sustainable Development Goal 3.3—ending epidemics of malaria and other communicable diseases.

Objectives

Primary objective

• •To identify the environmental and household determinants associated with increased mosquito presence in the urban field practice area of Mangalore.

Secondary objectives

• •To describe household-level preventive practices and community perceptions of mosquito-borne diseases.
• •To estimate the prevalence of self-reported mosquito-borne disease in the last 12 months and assess associations with environmental and behavioral factors after adjustment.

Methodology

This community-based cross-sectional study was conducted in Mangalore, a coastal city on the western coast of Karnataka, a South Indian state. Mangalore with an area of 132.4 km ² is situated between 12°50′30″ N to 13°01′00″ N and 74°48′0″ E to 74°55′00″ E coordinates, is a tropical river basin, and has a humid climate of peninsular India. Mangalore is bounded by the western Ghats to the east, the Arabian Sea to the west, Kerala to the south, and the Udupi district to the north. Mangalore, the district headquarters of Dakshina Kannada, is administered by a city corporation founded in 1865 and consists of 60 wards. ³⁰ Wards 27, 28, 31, 32 and 33 were chosen as study areas ( Figure 1). According to the Census of India 2011, the population of the Mangalore City Corporation was 499,487. In the absence of a more recent official census, population projections indicate that it may have grown to approximately 700,000 residents by 2024. ³¹

Figure 1. Map showing the Mangalore City Corporation study areas consisting of 60 wards.

The study was conducted between September and October 2024, the late monsoon transition season in Mangalore. The sample size was calculated based on a previous study conducted in Mangalore, Karnataka, ³² which reported that 83% of the people used preventive measures such as mosquito nets to prevent mosquito bites, using this as our anticipated proportion and 10% relative precision, 97.5% quantile of the standard normal distribution, and considering 20% non-response rate as 95 sample size was calculated as follows:

where p = 83%, d = 10% of 83% = 8.3%, Z = 1.96

$$n=\frac{Z_{1-\frac{a}{2}}^{2}P(1-P)}{d^2}$$

$$n=\frac{{1.96}^{2}0.83(1-0.83)}{{0.083}^2}$$

$$n=\frac{0.5421}{0.006889}=78$$

To account for 20% of the non-responses, 95 households were selected.

The study protocol was approved by the Institutional Ethics Committee, Kasturba Medical College Mangalore with No: IEC KMC MLR 09/2024/587, followed by permission from the Head of the Institute. The written consent was taken on an ‘Informed consent form’ which was provided to the participants ≥18 years of age for signature, and the participants <18 years of age were excluded from the study. All participants had the right to withdraw at any stage of the study, and all incomplete responses were considered withdrawal and excluded from the analysis. The study participants were approached from a household that was selected from five wards out of the total 60 wards present under the Mangalore City Corporation on the basis of the feasibility and representativeness of urban residential areas; one reliable informant residing in the household for more than at least 1 year who was aware of the household conditions and consented was included in the study, whereas temporary residents of less than a year, apartment complexes, and households without adult personnel were excluded. Apartment complexes were excluded because shared environmental exposures (e.g., corridors, rooftop tanks, and common waste areas) are not uniformly perceived by all residents, and this misconception may increase the risk of misclassification. Stand-alone houses, with clearer and self-contained environmental boundaries, allow for more consistent and valid household-level data collection.

The number of households included in our study was divided equally among each ward; that is, a total of 19 households from each of the selected wards were considered for the study. A random street was selected from each ward, and then, standing at the end of each randomly selected street, a random starting house was identified by selecting one house from the first five houses on that street. From this starting point, every 3 ^rd house present on the left side of the lane was considered for the study, which was continued in a clockwise order; in case of failure to meet the inclusion criteria, absence, or denial to take part in the study, the next house was considered. When the head of the household was absent during the time of data collection, information was collected from the oldest adult residing in the household.

A pretested, semi-structured, and internally validated questionnaire was developed for data collection. The questionnaire was constructed following an extensive literature review using electronic databases such as PubMed, Scopus, and Google Scholar, covering studies published between 2000 and 2024 that examined mosquito ecology, vector proliferation and nuisance, community perceptions, preventive practices and self-reported mosquito-borne illnesses in urban settings. Relevant national guidelines from the National Vector Borne Disease Control Programme (NVBDCP) were also reviewed to ensure contextual alignment. ^{33, 34}

Content validity was assessed by a panel of three public health specialists and two medico social workers—from the Department of Community Medicine, Kasturba Medical College, Mangalore. Each expert independently assessed the questionnaire items for clarity, relevance, simplicity, and ambiguity using a four-point Likert scale (1 = not relevant, 4 = highly relevant). The item-level content validity index (I-CVI) was calculated as the proportion of experts rating each item as either 3 or 4. Items with an I-CVI ≤ 0.78 were revised or removed on the basis of the panel’s feedback. The scale-level content validity index (S-CVI/Ave), computed as the average of all the I-CVIs, score ≥ 0.8 was considered satisfactory, indicating good overall content validity. The questionnaire was pilot tested among 20 household participants from the target population to ensure clarity and feasibility. Feedback from the pilot led to minor revisions in wording and the order of items to enhance clarity and comprehension. Internal consistency for each construct was assessed using Cronbach’s alpha, with values ≥0.7 considered acceptable, indicating reliable measurement of the respective domains.

The data were recorded after informed consent was obtained from the head of households. The questionnaire was self-administered; however, for illiterate participants, it was administered by the investigators. The questionnaire was designed to assess community perceptions of mosquito nuisance and other determinants. Operational definitions for key study variables were established prior to data collection. Mosquito presence was defined as the observed occurrence of adult mosquitoes within or around the household at the time of the survey and was further quantified by the frequency of sightings reported by the informant.

Accordingly, the nuisance score reflects the participant’s perceived intensity of mosquito activity compared with the previous year, with “mild” indicating that mosquitoes were rarely noticed or caused minor annoyance (≤2 bites per day on average), and “moderate-to-severe” indicating that mosquitoes were frequently noticed, causing significant annoyance or ≥3 bites per day on average, interfering with daily activities or sleep. ^{21 , 22} Self-reported mosquito-borne disease was operationally defined as any episode of malaria, dengue, chikungunya, or other vector-borne illness experienced in the past 12 months, as recalled by the participant. Only episodes for which the participant sought treatment or received a probable diagnosis from a healthcare professional were considered to improve the reliability of reporting. Water stagnation was defined as visible standing water persisting for more than 48 hours after rainfall or resulting from blocked drains within the household compound or immediate neighborhood.

The data collected were entered into MS Excel and analyzed using Jamovi version 2.6.26. Descriptive statistics are presented as frequencies and proportions. Confidence intervals (95%) for proportions were computed using the exact binomial method. The association between two categorical variables was assessed using the chi-square test. The strength of the association between the independent and dependent variables were measured using odds ratios (OR) with corresponding 95% confidence intervals (CI) and significant p value of <0.05. Multivariate logistic regression models were used to determine the independent predictors of mosquito presence, preventive practices within households, and the occurrence of mosquito-borne diseases. Variables with p value <0.2 in the univariate logistic regression were included in the multivariate model. To adjust for familywise error rate (FWER), Holms correction was applied to the p values.

Results

The study included 95 households; the majority of the participants were above the age of 45 years (72.6%), and most were females (70.5%). The majority of participants (94.8%) were educated and 61.1% reported having access to a healthcare facility within 2 km of their residence ( Table 1).

Table 1. Socio-demographic characteristics of study participants (N = 95).

Socio-demographic details	n (%)	95% CI
Age (in years)
18-30	12 (12.6)	6.70% - 21.03%
31-45	14 (14.7)	8.30% - 23.50%
46-60	34 (35.8)	26.21% - 46.30%
>60	35 (36.8)	27.20% - 47.40%
Gender
Male	28 (29.5)	20.60% - 39.71%
Female	67 (70.5)	60.30% - 79.44%
Education
Illiterate	5 (5.2)	1.73% - 11.90%
Primary School	11 (11.6)	5.92% - 19.80%
High School + PUC	46 (48.4)	38.04% - 58.90%
Degree	33 (34.7)	25.30% - 45.20%
Proximity of Nearby Health Care facility (in km)
0-2	58 (61.1)	50.50% - 70.90%
>2-4	25 (26.3)	17.81% - 36.40%
>4	12 (12.6)	6.70% - 21.03%

Among the study participants, 42.1% reported an increase in mosquito breeding sites past year. Most participants perceived a moderate to severe mosquito presence both indoors (67.4%) and outdoors (87.4%). The evening hours (4–8 pm) were identified as the peak biting time by 71.6% of the respondents. Mosquito nuisance was most prominent during the rainy season (69.5%), and 30.5% reported disturbed sleep or discomfort due to mosquito bites at night ( Table 2).

Table 2. Perceived mosquito proliferation and nuisance (N = 95).

Perception	n (%)	95% CI
Increase in mosquito breeding sites in last 1 year (yes)	40 (42.1)	32.04% - 52.70%
How would you rate the presence of mosquitoes inside the house?
Mild	31 (32.6)	23.40% - 43.02%
Moderate to Severe	64 (67.4)	57.0% - 76.64%
How would you rate the presence of mosquitoes outside the house?
Mild	12 (12.6)	6.70% - 21.03%
Moderate to Severe	83 (87.4)	79.00% - 93.0%
Time of peak mosquito biting
Morning (6-10 am)	8 (8.4)	3.71% - 15.92%
Evening (4-8 pm)	68 (71.6)	61.40% - 80.40%
Night (After 8 pm)	19 (20.0)	12.50% - 29.50%
Mosquito bites causing disturbed sleep or discomfort at night	29 (30.5)	21.50% - 40.82%
Weather conditions in which mosquitoes are mostly prevalent
Rainy season	66 (69.5)	59.20% - 78.51%
Hot weather	16 (16.8)	9.94% - 25.90%
No noticeable difference	13 (13.7)	7.50% - 22.30%
Overall mosquito nuisance in locality compared to past one year
Mild	38 (40.0)	30.10% - 50.60%
Moderate to Severe	57 (60.0)	49.44% - 69.92%

Water stagnation 74.7% and dense vegetation 54.7% were the most frequently reported environmental conditions associated with mosquito proliferation. Other contributing factors included flowerpots (84.5% among those reporting stagnation), nearby water bodies (20%) and garbage dumping (13.8%) ( Table 3).

Table 3. Determinants of mosquito breeding (N = 95).

Determinants of mosquito breeding *	n (%)	95% CI
Water stagnation (n = 71)	71 (74.7)	64.80% - 83.10%
Puddles	53 (74.6)	62.92% - 84.23%
Flowerpots	60 (84.5)	74.0% - 92.0%
Construction sites	23 (32.4)	21.80% - 44.10%
Garbage Dumping	13 (13.7)	7.50% - 22.30%
Presence of water body	19 (20.0)	12.50% - 29.50%
Presence of dense vegetation	52 (54.7)	44.20% - 65.0%
Water storage in uncovered containers	12 (12.6)	6.70% - 21.03%
Presence of water leaks or overflow from pipes and tanks	8 (8.4)	3.71% - 15.92%

^* Multiple responses.

Most households reported using chemical measures such as sprays, vaporizers, or coils (91.6%), closing doors and windows (81.1%), and using mosquito nets or screens (55.8%). Approximately half (51.6%) cleaned their surroundings daily, whereas 38.9% checked for water stagnation weekly. However, 58.9% stated that the municipality rarely conducts cleaning activities, and 50.5% reported no anti-mosquito fogging by local authorities ( Table 4).

Table 4. Behavioural and preventive measures by participants for mosquito control (N = 95).

Preventive measures		n (%)	95% CI
Type of measure *
Chemical		87 (91.6)	84.10% - 96.30%
Mosquito nets, window screens or meshes		53 (55.8)	45.23% - 66.0%
Closing doors and windows		77 (81.1)	71.72% - 88.40%
Other personal measures		15 (15.8)	9.12% - 24.70%
Mosquito repellent/coils
Yes, daily		24 (25.3)	16.91% - 35.22%
Yes, occasionally		23 (24.2)	16.01% - 34.10%
Never		48 (50.5)	40.10% - 61.0%
Insect repellent before sleep
Yes		12 (12.6)	6.70% - 21.03%
No		83 (87.4)	79.0% - 93.30%
Cleaning of surroundings	By household members
	Daily	49 (51.6)	41.10% - 62.0%
	Weekly	31 (32.6)	23.40% - 43.02%
	Occasionally	15 (15.8)	9.12% - 24.70%
	By municipality
	Yes	39 (41.0)	31.10% - 51.62%
	No	56 (58.9)	48.40% - 68.94%
Water stagnation (Checking and eliminating stagnation)
Daily		17 (17.8)	10.80% - 27.10%
Weekly		37 (38.9)	29.11% - 49.60%
Occasionally		30 (31.6)	22.42% - 41.92%
Never		11 (11.6)	5.92% - 19.80%
Check holes in window screens/mosquito nets
Regularly (monthly/more)		24 (25.2)	16.91% - 35.22%
Occasionally		17 (17.9)	10.80% - 27.10%
Rarely		17 (17.9)	10.80% - 27.10%
Never		37 (39.0)	29.11% - 49.50%
Anti-mosquito fogging by local authorities
Frequently #		6 (6.3)	2.40% - 13.24%
Occasionally		41 (43.1)	33.03% - 53.72%
Never		48 (50.5)	40.10% - 61.0%

^*Multiple responses.

^# Frequently: weekly or monthly.

Awareness of mosquito-borne diseases was high (92.6%) reporting knowledge of at least one mosquito-borne disease. Malaria (96.6%) and dengue (93.2%) being the most commonly recognized diseases. Most participants (96.8%) believed that education and awareness campaigns are crucial for disease prevention. A majority (83.1%) identified both clean and dirty water as potential breeding sources, and 81.2% noted that mosquito proliferation peaks during the rainy season ( Table 5).

Table 5. Awareness and perception of mosquito borne diseases and mosquito control measures (N = 95).

Awareness and perception	n (%)	95% CI
Awareness about mosquito borne disease *	88 (92.6)	85.41% - 97.0%
Malaria	85 (96.6)	90.40% - 99.3%
Dengue	82 (93.2)	85.8% - 97.5%
Chikungunya	29 (33)	33.0% - 43.8%
Zika	8 (9.1)	4.01% - 17.13%
Filariasis	9 (10.2)	4.80% - 18.53%
Perception of Preventive measures to be taken to avoid Mosquito-borne diseases *
Regular cleaning of surroundings	76 (80.0)	70.54% - 87.51%
Eliminating stagnant water	71 (74.7)	64.80% - 83.10%
Usage of insecticide sprays	50 (52.6)	42.12% - 63.0%
Usage of mosquito nets	46 (48.4)	38.04% - 68.90%
Education and spreading awareness	92 (96.8)	91.10% - 99.34%
Perception of community on determinants of mosquito proliferation
a) Water stagnation
Clean water	16 (16.8)	9.94% - 25.90%
Dirty water	44 (46.3)	36.02% - 56.84%
Both clean and dirty water	35 (36.8)	27.20% - 47.40%
b) Seasonal variation	85 (89.5)	81.50% - 94.84%
Rainy season	69 (81.2)	71.24% - 88.84%
Summer season (Dry Hot weather)	16 (18.8)	11.20% - 28.80%

^*Multiple responses.

In the past year, 29.4% of the participants reported suffering from a mosquito-borne disease, primarily dengue (71.4%) and malaria (28.5%). Common symptoms included fever (96.4%), headache (78.5%), joint pain (60.7%), and muscle pain (46.4%). Most sought treatment at private clinics (82.2%). Post recovery complications such as generalized weakness were reported by 92.3% of the patients (N=13) ( Table 6).

Table 6. Self-reported cases and outcomes of mosquito borne diseases (N=95).

Self-Reported cases	N (%)	95% CI
Suffered from any mosquito-borne disease in the past 1 Year
Yes	28 (29.4)	20.60% - 39.71%
Mosquito borne Disease (N = 28)
Malaria	8 (28.5)	13.22% - 48.70%
Dengue	20 (71.4)	51.33% - 86.80%
Symptoms * (N = 28)
Fever	27 (96.4)	81.70% - 99.91%
Headache	22 (78.5)	59.10% - 91.70%
Joint Pain	17 (60.7)	40.60% - 78.50%
Rash	2 (7.1)	1.8% - 23.50%
Muscle Pain	13 (46.4)	27.51% - 66.13%
Others	17 (60.7)	40.60% - 78.50%
Place of Treatment (N = 28)
Public Health Centre	5 (17.8)	6.1% - 36.9%
Private Clinic	23 (82.2)	63.11% - 93.94%
Complications after recovery * (N = 13)
Weaknesses	12 (92.3)	64.0% - 99.8%
Cold	1 (7.6)	0.20% - 36.03%
Leg Pain	1 (7.6)	0.20% - 36.03%
Headache	1 (7.6)	0.20% - 36.03%
Eye Pain	1 (7.6)	0.20% - 36.03%

^*Multiple choice question.

Variables with p value <0.2 in the univariate logistic regression were included in the multivariable logistic regression model and both unadjusted and adjusted odds ratios (ORs) with 95% confidence intervals were presented. After applying Holm’s correction for the familywise error rate (FWER), the logistic regression model showed no significant associations between mosquito density and environmental factors such as water stagnation, construction sites, nearby water bodies, dense vegetation, garbage dumping, type of water storage, or water leaks ( Table 7).

Table 7. Associations of environmental factors with mosquito density: univariate and multivariate logistic regression analysis (N=95).

Variable	Increased mosquito density		Unadjusted OR (95% CI)	p value	Adjusted OR (95% CI)	p value
	Yes (%) n = 49	No (%) n = 46
Presence of water stagnation Yes No	34 (69.3) 15 (30.6)	22 (47.8) 24 (52.1)	2.5 (1.1-5.7)	0.198 #	2.1 (0.9-5)	0.097
Presence of construction sites Yes No	17 (34.6) 32 (65.3)	6 (13.0) 40 (86.9)	3.5 (1.2-10.0)	0.098 #	3 (1-8.8)	0.084
Presence of any water body (pond, lake, etc.) Yes No	8 (16.3) 41 (83.6)	7 (15.2) 39 (84.7)	1.1 (0.4-3.3)	0.882	-	-
Presence of Dense Vegetation Yes No	31 (63.2) 18 (36.7)	21 (45.6) 25 (54.3)	2.1 (0.9-4.6)	0.425	-	-
Presence of Garbage dumping sites Yes No	8 (16.3) 41 (83.6)	5 (10.8) 41 (89.1)	1.6 (0.5-5.3)	0.878	-	-
Type of Water storage Covered containers Uncovered containers	41 (83.6) 8 (16.3)	42 (91.3) 4 (8.6)	2.1 (0.6-7.3)	0.789	-	-
Presence of water leaks or overflow from tanks or taps Yes No	6 (12.2) 43 (87.7)	2 (4.3) 44 (95.6)	3.1 (0.6-16.1)	0.664	-	-

OR = Odds Ratio; CI = Confidence Interval.

^# Holm-adjusted p ≤ 0.2 included in multivariable model.

The use of mosquito repellents or coils was positively associated with self-reported mosquito-borne disease (OR = 3.7; 95% CI: 1.4–9.6; p = 0.024), although this finding likely reflects greater repellent use in households experiencing greater mosquito burden rather than indicating any causal influence of repellents on disease risk. Other preventive practices including the use of mosquito nets, insecticide sprays, or the regular repair of screens showed no statistically significant associations with disease occurrence ( Table 8).

Table 8. Association between the mosquito borne disease prevalence and preventive measures taken inside the house (N = 95).

Variable	Presence of mosquito-borne disease		OR (95% CI)	p Value
	Yes (%) n = 28	No (%) n = 67
Repair or check holes in window screens or mosquito nets Yes No	9 (32.1) 19 (67.8)	32 (47.7) 35 (52.2)	0.5 (0.2-1.2)	0.483
Use of mosquito repellents or coils Yes No	20 (71.4) 8 (28.5)	27 (40.3) 48 (59.7)	3.7 (1.4-9.6)	0.024 #
Use of electric mosquito bats or insecticide sprays Often Rarely	11 (39.2) 17 (60.7)	22 (32.8) 45 (67.1)	1.3 (0.5-3.3)	>0.999
Usage of Mosquito nets at night Yes No	14 (50.0) 14 (50.0)	37 (55.2) 30 (44.7)	0.8 (0.3-1.9)	>0.999

OR = Odds Ratio; CI = Confidence Interval.

^# Statistically significant (Holm-adjusted p ≤ 0.05).

Discussion

The present study, which was conducted within the Mangalore City Corporation, provides valuable insights into the determinants of mosquito population density and community responses in the urban setting of coastal Karnataka. These findings affirm that urban mosquito breeding and disease transmission are influenced by a complex interplay of environmental, behavioral, and infrastructural factors.

In the present study, a majority (95%) had some level of education and were above 45 years of age, with a predominance of female respondents, and 61% reported nearby healthcare facilities within 2 km, indicating relatively good access to healthcare in the study area. These findings match those of prior studies highlighting that educational level and proximity to health services may influence awareness and preventive behaviors related to vector-borne diseases, although not necessarily with consistent environmental control practices. ^{27, 28}

More than two-fifths of the respondents perceived an increase in mosquito breeding sites within the last year and 60% of participants reported experiencing moderate to severe mosquito nuisance in their locality over the past year, which was consistent with urbanization-related ecological changes in the study setting. This aligns with the national trends of urban vector expansion reported in the National Vector Borne Disease Control Programme (NVBDCP) surveillance data. ³⁴ Notably, 71.5% of the participants indicated that evenings were the peak time for mosquito activity, and 30.5% reported that mosquito bites disrupted their sleep, which is consistent with Aedes aegypti’s day-biting habits, whereas Anopheline mosquitoes are known for nocturnal activity, which suggests the presence of mixed mosquito species in the study area. ²²

Additionally, 67.4% of the participants described a moderate to severe presence of mosquitoes indoors, whereas 87.3% reported similar conditions outdoors. This finding is comparable to data from Kampango et al., who reported an average of 85.93% An. gambiae s.l. bites per night, with 66% occurring indoors and 34% outdoors, peaking between 22:00 and 03:00 in a rural community in Chókwè District, southern Mozambique. ³⁵

In terms of environmental factors, 69.4% of the participants reported high mosquito activity during the rainy season, which corresponds to the behavioral patterns of Aedes aegypti and Culex quinquefasciatus, both of which are well adapted to peridomestic environments. ¹ This aligns with a cross-sectional study by Mahgoub et al. in Barakat and El-Kareiba, Sudan, which reported a high number of positive habitats during the rainy season, whereas the lowest numbers were reported during the hot season followed by the dry season, corroborating findings that climate and seasonal variation significantly influence mosquito density and disease transmission. ^{26, 36}

The primary breeding sites identified in our study included water stagnation (74.7%), dense vegetation (54.7%), and nearby water bodies (20%), which is consistent with previous studies demonstrating the importance of stagnant water and vegetation in vector proliferation. ^{21 – 23} Urban infrastructure projects, especially when poorly managed, contribute to temporary water stagnation, while dense vegetation offers resting sites and microclimatic conditions favourable to adult mosquitoes.

After applying Holm’s correction for the familywise error rate (FWER), the logistic regression model revealed no significant associations between mosquito density and environmental factors such as water stagnation (OR 2.5, 95% CI: 1.1, 5.7, p value – 0.198), construction sites (OR 3.5, 95% CI: 1.2, 10.0, p value – 0.098), nearby water bodies (OR 1.1, 95% CI: 0.4, 3.3, p value – 0.882), dense vegetation (OR 2.1, 95% CI: 0.9, 4.6, p value – 0.425), garbage dumping (OR 1.6, 95% CI: 0.5, 5.3, p value – 0.878), type of water storage (OR 2.1, 95% CI: 0.6, 7.3, p value – 0.789), or water leaks (OR 3.1, 95% CI: 0.6, 16.1, p value – 0.664). Water stagnation near residential compounds may often result from inadequate municipal drainage systems—a structural determinant beyond individual household control. ^{25, 26}

Preventive measures among participants were notable, as 91.5% used chemical repellents (insect sprays, vaporizers, and smoke), 81% kept doors and windows closed, 55.7% utilized physical barriers such as mosquito nets or screens, and 15.7% employed other personal measures. These findings are consistent with household-level studies from Mumbai and Sri Lanka, which reported that city respondents, on the other hand, were more likely to use liquid repellents and mosquito sprays, perhaps owing to their ease of use and their immediate, visible effects and commercial availability. ^{37, 38}

In terms of community cleaning practices, 51.5% of individuals reported cleaning their surroundings daily; additionally, 38.9% checked for water stagnation weekly; in contrast, 84.6% never reported water stagnation to authorities. Unfortunately, 58.9% indicated that municipal cleaning was infrequent in the locality, reflecting gaps in sustained vector management. Notably, 25.2% of the participants regularly checked window screens or mosquito nets for damage, and 50.5% stated that there had been no anti-fogging efforts by local authorities, underscoring the need for better municipal participation and routine surveillance-based larval control, as emphasized in the National Framework for Malaria Elimination 2016–2030. ³³ This is in contrast with Mahalakshmi et al.’s findings in northern Gujarat, where 88% cleaned their homes daily, 57.3% cleaned their surroundings weekly, and 82% actively avoided water stagnation. ³⁹

High awareness of mosquito-borne diseases (92.6%), particularly malaria and dengue, is comparable to that reported in other urban studies. However, awareness alone does not guarantee effective preventive action, a well-documented paradox in vector control studies. For example, a community-based survey from Puducherry revealed that although 85.5% of the total respondents had heard of dengue fever and that most of them (82.7%) were aware that it is transmitted through mosquito bites, only 25.1% of participants were aware that the dengue mosquito breeds in clean water-holding containers. ⁴⁰

Despite these measures, approximately 30% of the participants reported suffering from a mosquito-borne disease in the past year, with dengue accounting for 71.4% of the cases and malaria accounting for (28.6%). Dengue virus consists of four antigenically distinct serotypes (DENV-1 to DENV-4), and the cocirculation of multiple serotypes such as DENV-1, DENV-2, and DENV-3 contributes to complex transmission patterns. Primary infection confers only short-term cross-protection, and secondary infection with a heterologous serotype is well known to carry a markedly increased risk of severe disease due to mechanisms such as antibody-dependent enhancement. Consequently, such clinically overt secondary infections are more likely to be detected and reported, which may explain the higher proportion of dengue cases observed in the study setting. ⁴¹

Common symptoms included fever (96.4%), headache (78.5%), joint pain (60.7%), and muscle pain (46.4%), which aligns with the WHO’s case definition for dengue and aligns with findings from a study conducted by Kumar et al. in a tertiary hospital in the Udupi district, which indicated that 83.9% of cases were due to dengue fever, presenting symptoms such as fever (99.1%), myalgia (64.6%), vomiting (47.6%), headache (47.6%), abdominal pain (37.5%), and breathlessness (17.8%). ^{9, 42}

In this study, 92.3% of the participants reported experiencing complications post recovery, suggesting prolonged morbidity, an often underrecognized component of dengue disease burden. This contrasts with a study in Vietnam by Tam et al., who reported that 12.5% of participants experienced alopecia, 11.1% had blurred vision, 9.5% faced concentration difficulties, and 8.5% suffered from fatigue following dengue infection. ⁴³

The present study revealed that 82.1% of the population perceives water stagnation as a significant factor in mosquito proliferation, indicating that it is a major factor in the prevalence of mosquito-borne diseases. Construction sites, water bodies, dense vegetation, and water storage in open containers were identified as significant determinants. Additionally, 17.8% of individuals perceive water leakage from tanks and taps as key determinants, indicating that water leakage from household stores is a significant factor in the prevalence of mosquito-borne diseases. Overall, these factors contribute to the prevalence of mosquito-borne diseases.

Interestingly, the use of mosquito repellents or coils was significantly associated with a greater incidence of mosquito-borne diseases (OR 3.7, 95% CI: 1.4, 9.6, p = 0.024) suggesting a reactive response: — households in high-risk areas or with prior illness episodes are more likely to adopt repellents. This phenomenon has been described in behavioral epidemiology as the “reverse causation effect”. ⁴⁴ This finding also suggests possible improper usage or overreliance on repellents without addressing environmental breeding sites. Similar gaps between knowledge and action have been reported in prior studies. ²⁷

Strengths

The key strength of this study is its systematic, community-based sampling across selected wards using a random start method for street selection and a structured household sampling approach. Standardized data collection, operational definitions, and a pilot-tested data collection tool enhance reliability. The study also examined both environmental and behavioral determinants, allowing for a more holistic assessment.

Limitations

As a cross-sectional study, temporal causality between determinants and mosquito density could not be established. Entomological indices (e.g., the Breteau or House index) were not measured, limiting direct quantification of vector density. Self-reported disease history may be subject to recall bias, although the one-year recall period likely limits its magnitude. This study was conducted during the late monsoon–transition period in Mangalore, when the number of breeding sites typically begins to decline, which may influence the observed mosquito density compared with that in the peak monsoon months. This study did not assess local insecticide resistance patterns, which is relevant given the presence of Anopheles stephensi in the region. Additionally, behavioural practices and environmental exposures were based on self-reports and may be influenced by social desirability or misclassification, although the structured questionnaire and short recall frames were designed to minimize such biases. Nonetheless, this study provides strong evidence linking environmental conditions and behavioral factors to perceived mosquito proliferation, offering valuable insights for targeted urban vector-control strategies.

Conclusion

In conclusion, this study provides an integrated approach by identifying key environmental determinants of mosquito presence while simultaneously evaluating community preventive measures, perceptions and self-reported mosquito-borne disease burden, offering context specific insights from Mangalore that complement previous studies in other settings. While knowledge and attitudes in the community were generally adequate, their association with actual preventive practices was modest, indicating a persistent gap between awareness and action. However, proper intervention by local authority is necessary to combat the main environmental factors responsible for mosquito breeding. This highlights the gaps found in our study, where despite widespread awareness of mosquito-borne diseases, respondents acknowledged limited knowledge of effective preventive measures and exhibited suboptimal preventive practices. Hence, in addition to awareness, there is a dire need to provide the right personnel and services to combat mosquito-borne diseases.

Recommendation

Awareness and campaigns

Nearly one hundred percent of the participants believed that spreading awareness of mosquito-borne diseases and mosquito control measures would help reduce the incidence of mosquito borne diseases in the community. This can be accomplished through various means, such as public service announcements, social media campaigns, and community outreach programs.

Elimination of water stagnation

Local authorities and communities must proactively repair damaged roads, cover open drains, strengthen vector control at construction sites, ensure proper waste disposal to eliminate standing water, prevent mosquito breeding, and reduce the risk of vector-borne diseases.

Clearing up of dense vegetation

Dense vegetation was identified as a significant determinant of mosquito presence in household neighborhoods. Local authorities should ensure regular clearing of vegetation and maintenance of cleanliness in these areas.

Further research

Further research to identify the factors leading to an increase in the mosquito population and community and local government measures can help identify additional intervention strategies. This can involve qualitative research methods, such as in-depth interviews, field research, or focus groups, to gain a deeper understanding of the underlying issue.

Policy implications

This study underscores the need for stronger municipal policies focused on eliminating water stagnation, regulating construction sites, and maintaining vegetation to address key environmental drivers of mosquito proliferation. Sustained community-level awareness campaigns are essential to improve preventive practices and encourage early health-seeking behavior. Additionally, regular operational research and entomological surveillance should be integrated into local vector control programs to ensure data-driven, targeted interventions.

Ethical approval

The study protocol was approved by the Institutional Ethics Committee, Kasturba Medical College Mangalore with No: IEC KMC MLR 09/2024/587, followed by permission from the Head of the Institute. The written consent was taken on an ‘Informed consent form’ which was provided to the participants ≥18 years of age for signature, and the participants <18 years of age were excluded from the study. All participants had the right to withdraw at any stage of the study, and all incomplete responses were considered withdrawn and excluded from the analysis.

Funding Statement

The author(s) declared that no grants were involved in supporting this work.

Data availability

Underlying data

The data set used and/or analyzed during the current study are available from the online repository (figshare) DOI- https://doi.org/10.6084/m9.figshare.29144975 ⁴⁵ (Tables 1–8).

Extended data

The data include participant information sheet, Informed consent form and questionnaire are available from the online repository (figshare) DOI- https://doi.org/10.6084/m9.figshare.29144990. ⁴⁶

Data are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 Public domain dedication).