Prevalence and Associated Factors of Dengue Virus Circulation in the Rural Community, Handeni District in Tanga, Tanzania

Dengue virus is among the most important re-emerging arbovirus that causes global public health attention. Dengue has historically been thought of as an urban disease that frequently occurs in rapidly urbanized settings. However, dengue has become more widespread in rural regions in recent years. Understanding the changing dengue epidemiology in different geographical settings is important for targeted intervention. In Tanzania, dengue fever is not frequently reported because of the poor surveillance infrastructure, underestimation, and a lack of consideration of dengue as a priority. Therefore, the true burden as well as the risk factors for increased transmission has not been fully ascertained, particularly in rural areas. A cross-sectional community-based study was conducted in June 2021, involving a total of 362 participants of all age groups. We investigated the prevalence of acute dengue infection, seroprevalence, and associated factors among the community in three villages of the rural Handeni district. The prevalence of acute dengue infection (based on PCR) was 2.2% (8/362). Dengue-specific IgM and IgG antibodies were detected in 3.3% (12/362) and 5.2% (19/362) of the participants, respectively. Adult participants who were having vegetation around their houses were more likely to be DENV seropositive (AOR = 2.4, CI = 1.88–4.18, p value = 0.05). Children living in houses with garbage pit around their households were less likely to be DENV seropositive (AOR = 0.13, CI = 0.03–0.56, p value <0.01). DENV continues to circulate in rural Tanzania, causes an alarming situation, and necessitates prompt public health action to enhance vector surveillance and control in rural communities.


Introduction
Dengue virus (DENV) infection is a mosquito-borne viral disease that contributes to the global health challenge due to its high morbidity and mortality with more than half of the world's population at risk of infection [1,2].Each year, up to 400 million people are infected with DENV, and 40000 die from severe DENV infection, while over 80% of infected individuals are having mild symptoms or are asymptomatic [2,3].Dengue is caused by four DENV serotypes, DENV 1-4.Dengue can manifest with a wide range of clinical signs, from mild febrile illness to severe plasma leakage that can result in life-threatening shock [4].Te infammatory cytokine storm and the host response to DENV infection cause the endothelial cells to become more permeable, which causes vascular leakage and contributes to severe dengue [5].One DENV serotype during primary infection results in long-term immunity against that specifc serotype but short time immunity against another serotype, secondary infection with a diferent serotype increases the likelihood of severe disease [6].Te observed expansion and outbreaks around the globe are facilitated by rapid urbanization, increased travel, globalization, and climate change [7,8], which afects vector distribution and virus survival.
In Africa, DENV infection is underreported because of the poor surveillance infrastructure and diagnostic capacity, leading to under-estimation of the disease [9,10], in addition, DENV infection is not considered a priority in many African countries Ministry of Health, Tanzania included.Hence, the true burden of DENV infection is not known.Evidence suggests that DENV infection in African countries' populations may be more widespread than reported [11].More importantly, DENV infection shares endemicity and similar clinical presentation with chikungunya, malaria, typhoid, and infuenza [12], thus leading to difculty in diagnosis when only clinical diagnosis is used in ruling out the cause of infection.
Dengue vaccine, Denvaxia, is currently approved for use in children aged 9 to 16 who have previously had a laboratory-confrmed DENV infection and who live in dengueendemic areas.Tis vaccine is only approved in the United States.Tus, limiting control measures in poor resource countries like Tanzania.Due to the lack of a dengue vaccine or treatment options, treatment is supportive to alleviate the symptoms [13].
Since 2010, Tanzania has experienced reoccurring outbreaks of DENV infection, with the most recent one occurring in 2019 and infecting 6917 people, predominantly in Dar es Salaam and Tanga [14].Te DENV infection season in Tanzania mainly peaks from May to June [14].Studies conducted during the outbreak showed that the DENV virus primarily spreads in urban areas [15,16], where Aedes mosquitoes thrive because they prefer urban areas [17,18].Rapid urbanization is often associated with the emergence and spread of dengue diseases by creating favorable breeding sites such as discarded plastic containers and abandoned car tires [18], as compared to rural areas [19].More importantly, high population density may increase the risk of infection, although more accessible health systems increase the likelihood of an accurate diagnosis, as well as the availability of tests and more sensitive medical personnel as compared to rural areas.
Many studies on nonmalaria fever have been conducted in Tanzania, but few highlighted those patients with acute DENV infections are often misdiagnosed and more often treated with antimalarial or antibiotics [20][21][22][23].Besides, the consequences of misdiagnosis may lead to economic loss, the development of drug resistance of malaria/bacterial strains (due to over-prescribing of antimalarials and antibiotics), ongoing transmission, and the risk of increased morbidity and mortality.Evidence suggests that several factors are associated with DENV infection transmission, including vegetation, uncovered containers, animal husbandry in the peridomestic environment, stagnant water, and a humid/ warm environment favoring the breeding of mosquito vectors [18,24,25].But these studies have been conducted in urban areas thus limiting the epidemiology in the rural community in Tanzania.In other settings, the epidemiology of DENV infection is changing, with reports indicating a global rise in rural infections, particularly in Africa [10,26].Conversely, in Gabon, a study suggested minimal DENV infection circulation in rural areas [27], and in Kenya a study reported no evidence of DENV infection circulation in rural communities [28].
Te attempts for control are hampered by not understanding the full extent of the epidemiology of DENV infection, especially in rural areas, where it receives little attention.Tus, this study aimed to investigate the prevalence and associated factors and documents the clinical features of DENV infection among the community in a rural area of Tanzania.

Materials and Methods
2.1.Study Setting.Tis was a cross-sectional study conducted in the Bondo area, Handeni rural, Tanga region, Tanzania.Tanga is located at an elevation of 309 meters above sea level.Its coordinates are 5 °22′60″N and 38 °34′60″E in DMS (degrees, minutes, and seconds).Te area has an annual rainfall of more than 1,212 mm with monthly rainfall peaks in April and May (wet season) of over 470 mm, and September to October (short rains) with over 250 mm [29].Tanga is endemic to malaria with a prevalence of 14.6% [30], while the prevalence of fever is 14.8% [31], which could generally be attributed to viral, bacterial, or malaria infection.Most DENV transmission occurs during the rainy season as a result of increased Aedes vector abundance.Tree villages have been purposively selected, namely, Bondo, Kwamgwe, and Kwadoya.

Sample Size and Inclusion
Criteria.Simple random sampling was used to identify participants that met the inclusion criteria until the desired sample size was reached.Te minimum sample size for this study was calculated using the following formula n � z 2 pq/d 2 .In this equation, n is the sample size, z is the value of the standard normal distribution at the 5% level (1.96), p is the prevalence, q � 1 − p, and d is the precision level (0.05).Te prevalence of acute DENV infection is 38.2% [32].Sample size � 1.962 × 0.382 × (1 − 0.38)/(0.05) 2 .Te sample size was 362 for the detection of DENV infection.Te study included children (2-17 years old), and adults (18-70 years old) who volunteered to participate in the study, and who had lived in the area for at least a year.We excluded individuals with any signs of severe illness due to other causes other than arboviral diseases and those who were mentally unft.

Recruitment and Interview.
Community members and village leaders were involved before recruiting study participants.During recruiting and enrollment, randomly chosen participants were invited to the neighborhood dispensary or schools.Meetings were organized wherein study staf explained the purpose of the study and answered any questions in an open forum.Completing questionnaires with an interviewer started immediately after participants have consented to participate in the study.Participants were enrolled only after verifcation of the potential inclusion 2 Journal of Tropical Medicine criteria.After consenting, a structured Swahili questionnaire was used to collect information among the study population.
Te main contents of the questionnaire included general demographic characteristics (age, gender, marital status, education, average household (HH) income, and occupation).It also included social and personal life activities, such as travel history, outdoor sports, and any social events.Moreover, the questionnaire presented questions related to the prevention and control of mosquitoes.Also, the questionnaire included environmental sanitation (presence of garbage pits or vegetation near the house), housing characteristics, such as the building structure and roofng, and living conditions (average numbers of people sharing a room).Lastly, we collected information regarding the history of fever or illness for the past 24 hours and blood sample collection, and laboratory investigation (Supplementary File 1).

Blood Sample Collection and Serological Testing of DENV.
Approximately 0.4 ml of blood samples were drawn from the participants.
A few drops of blood samples were used for the detection of DENV-specifc IgM and IgG antibodies using on-site rapid dengue immunoglobulins M/G (IgM/IgG) (CTK Biotech, USA) according to the manufacturer's instructions.Briefy, blood samples were placed in a test kit, bufer was added, and waited for approximately 15 minutes before providing fnal results.Results interpretation for all rapid tests was done visually by the naked eye.In cases where the fnal results were ambiguous, a second confrmation from another laboratory technologist was requested.Two laboratory technologists who are accredited and licensed by the proper authorities conducted all of the testing.Te remaining blood samples were placed in EDTA tubes and stored temporarily in liquid nitrogen before transport to the Kilimanjaro Christian Medical University College (KCMUCo) laboratory for storage at −80 °C for further analysis.

RNA Extraction.
Viral RNAs were extracted from human whole blood using the Qiagen RNA Blood Mini Kits according to the manufacturer's instructions (Qiagen, Hilton Germany).Briefy, an aliquot of 30 µL of the blood sample was lysed and then homogenized using QIAshreader spin Coulmn.Te samples were applied to the spin column.Total RNA binds to the QIAamp membrane.Pure RNA was eluted in a total of 30 to 100 µL of RNA-free water and stored at −20 °C ready for PCR.

Detection of DENV by RT-PCR.
Te Aridia Zika/Dengue/CHIK real-time PCR test, in a one-step format, was used to detect the DENV infection (CTK Biotech, Inc., Poway, California, USA).Te lyophilized dengue-positive control was rehydrated in 100 μL of supplied PCR-grade water.A total of 5 μL of each extracted RNA, negative control, and positive control were pipetted in the respective wells in the plate.Te amplifcation reaction was performed in an AriaMx Agilent PCR machine (Agilent, Santa Clara, California, USA).Te thermocycler conditions started with reverse transcription at 45 °C for 15 minutes and an initial denaturation at 95 °C for 2 minutes, followed by 45 cycles of denaturation at 95 °C for 10 seconds and annealing/extension at 60 °C for 50 seconds.Fluorogenic data analysis of the samples and controls was performed by the real-time PCR thermocycler software, according to the manufacturer´s instructions.
2.7.Data Analysis.Te data collected were checked for completeness, cleaned, and entered into Statistical Package for Social Sciences (SPSS) version 26 for analysis.socioeconomic status (SES) proxy was created using principal component analysis (PCA) to investigate the independent efects of wealth indices on disease status [33].Categorical variables were summarized into frequency and percentage while the continuous variable was summarized by median and Interquartile range.Te chi-square (χ 2 )/Fischer exact test was used to determine the association between independent factors and the outcome variables.
To investigate the association between dengue virus circulation and independent factors, bivariate analysis was frst conducted for each potentially explanatory factor that was independently associated with DENV seropositivity.Variables with a p value <0.1, were considered in the multivariable analysis.Multivariable logistic regression was used to estimate the adjusted efect of factors associated with DENV.p values of ≤0.05 were recognized as signifcant.

Ethical Considerations.
Te approval was sought from the Kilimanjaro Christian Medical University College Research and Ethics Review Committee (CRERC) with certifcate number 2492 and the National Institute of Medical Research with certifcate number NIMR/HQ/R.8a/Vol.IX/ 3651.Permission to conduct this study was granted by the district medical ofcer (DMO) of the Handeni district and the village leaders.Written informed consent was obtained from all participants.Assent involved participants in the age range of 2 to 17 years, who were legally not able to consent by themselves.Parents or guardians consented on behalf of participants who were under 18 years.
Te participant who was not able to read or write an impartial witness signed on their behalf.Confdentiality was adhered to where participant information was available to the researcher only.Te study was conducted by following Good Clinical and Laboratory Practices.

Factors Associated with DENV Seropositivity among
Adults.In bivariate analysis, only wells/stagnant water and vegetation were signifcant predictors of DENV seropositivity (p set at <0.1) (Table 4).Te signifcant fndings of the bivariate analysis were further analyzed in the fnal model of multivariate analysis using logistic regression.Participants who had wells/stagnant water around their houses were less likely to be DENV seropositive (AOR � 0.01, CI � 0.02-0.08,p value <0.001).Participants who were having vegetation around their houses were more likely to be DENV seropositive (AOR � 2.4, CI � 1.88-4.18,p value � 0.05).

Factors Associated with DENV Seropositivity among
Children.In bivariate analysis, only average people sharing a room, garbage pit, vegetation, and roofng were signifcant predictors of DENV seropositivity (p set at <0.1) (Table 5).Te signifcant fndings of the bivariate analysis were further analyzed in the fnal model of multivariate analysis using logistic regression.Te fnal model of DENV seropositivity showed that two potential independent factors were signifcantly associated with DENV seropositivity: participants who had garbage pit around their houses were less likely to be DENV seropositive (AOR � 0.13, CI � 0.03-0.56,p value <0.01).Contrary to participants whose houses had grass or tile roofs, those with corrugated iron sheets had a higher likelihood of having DENV seropositive results (AOR � 3.90, CI � 0.94-16.11,p value � 0.05).

Discussion
Te study investigated the prevalence of acute DENV infection and seroprevalence as well as associated factors of DENV seropositivity among the community in the rural Handeni district in Tanga.Generally, we found that 3.3% and 5.2% of participants were seropositive to DENV IgM and IgG, respectively.Moreover, Bondo village had a higher IgM and IgG seroprevalence compared to other villages, although not signifcantly.Our fndings are lower than the study conducted in eight districts of Tanzania which reported a prevalence of 28.6% IgG-positive in the northeastern zone represented by Kilindi which is close to Kwamgwe [34].Lower results in our study could partly be attributed to the reason that our study was conducted in a community while the other study was conducted in hospital settings.Acute DENV infection was detected in 2.2% of participants, but there was no statistically signifcant diference between villages.Previous studies conducted in the same settings reported that none of the study participants were dengue-positive during the study period [22].Tis implies that DENV infection continues to emerge in recent years.Our fndings may justify the continued and persistent circulation of DENV infection in Tanzania.Also, our fndings explain the ongoing autochthonous transmission as evidenced by our data showing that all participants who tested positive for DENV infection did not travel outside the study area or region where they could contract the virus.Journal of Tropical Medicine

Journal of Tropical Medicine
Te study shows that there is no diference in DENV infection between villages, this suggests that the risk of being infected by DENV is relatively homogenous within the populations [35], and thus villages have relatively similar characteristics.It has been reported elsewhere that DENV infection has evolved from small outbreaks previously seen to large outbreaks experienced recently in 2019 [14,15].Tis implies that DENV infection is likely spreading and circulating to a larger extent and perhaps in other parts of Tanzania such as rural areas which goes unnoticed.Regarding the symptoms of DENV infection, the participant complained mainly about headache, abdominal pain, muscle pain, and arthralgia/tiredness. Te same observations were reported in a study conducted in Dar es Salaam which reported a higher number of DENV-infected patients who complained about headaches and muscle pain.One possible explanation for this discrepancy is because the study in Dar es Salam was conducted in a hospital setting [36].We conducted logistic regression diferently in adults and children.Our results showed that adult participants with the presence of wells/stagnant water around the home were less likely to be DENV seropositive.Tis may be explained by the fact that Aedes mosquitoes do not breed in wells/ stagnant water around houses.However, other studies in Tanzania reported that stagnant water is associated with DENV seropositivity [34].Further studies should be done to reveal this.
In the present study, adult participants who had vegetation around their HH were likely to be DENV seropositive compared to others.It has been reported that vegetation is associated with higher Aedes density as vegetation provides a more suitable resting habitat [37,38].It favors breeding sites close to dense vegetation including plantations which are linked to an increased risk of exposure for rural workers such as those in rubber and palm oil plantations, but it is also found to be established abundantly in urban areas [39,40].Previous studies conducted elsewhere displayed an association between greenery and the number of DENV infections with the perception that mosquitoes are abundant in highly vegetated areas [41,42].
Similar to adults, several factors have been linked to DENV seropositivity in children.Our study reports that sharing a room with more than 3 people per HH is associated with DENV seropositivity.Te same results have been long reported from studies conducted in Brazil [43] and Venezuela [44], indicating that more people in a room or crowded living is the most likely factor that increases DENV transmission.Tis may be relevant given the fact that Aedes mosquito bites during the day and the majority of parents do not cover their children with bednet when they sleep during the day.Participants with garbage pits were less likely to be DENV seropositive compared to those who did not.A diferent scenario has been reported in an urban area whereby areas with a high population density, presence of garbage pits, or solid waste disposal facilities are poor, and the open container may provide conducive breeding sites for Aedes mosquitoes [34,45].In contrast to the urban area, the reason for our observed results could be in the rural areas the frequent disposal of HH plastic wastes or containers is uncommon.In our study, participants' houses with corrugated iron sheets were more likely to be DENV seropositive compared to those with grass or tiles.Tis could be explained by the fact that houses with corrugated iron sheet tend to be warmer than other houses, thus afecting mosquito incubation and speeding up the transmission of the virus [46].

Strength and Limitation.
Te community-based study was more representative of the general population compared to other studies, which included febrile patients seeking healthcare.Te study could not establish a causal relationship between independent and dependent variables because of the cross-sectional nature of the study.A cohort study is recommended.In clinical practice, it is difcult to directly use IgM-based results to make a clinical decision.Detection of IgM antibodies in a patient is often deciphered as an indicator of recent and sometimes acute infections, however, caution should be considered due to the reason that false-positive IgM results are common, as a result of cross-reactivity with IgM antibodies to other, closely related arbovirus such as yellow fever and West Nile [47,48].IgM antibodies may remain detectable up to 3 months and more [49].However, the rapid test used in this study has a sensitivity and specifcity of 96.6% and 98.1%, respectively [50].

Conclusions
DENV continues to circulate in rural Tanzania and causes an alarming situation and there is an urgent need to develop targeted interventions and strengthen the surveillance system to address the issue.Te identifed factors of DENV seropositivity in this study could contribute to directing more targeted integrated preventive and control measures.In addition, community engagement is paramount in vector control.Tese fndings will help the Tanzania Ministry of Health plan for efective integrated control programs in rural communities.

Table 2 :
Seroprevalence and prevalence of DENV infection in the participants (N � 362).
* Calculated based on the wealth asset index.SES: socioeconomic status.