**4. Discussion**

We found that participation in outdoor sports activities and poor indoor daylight quality significantly increased the probability of contracting DF by 1.80- and 2.27-fold, respectively, in Guangdong Province. Our results also suggested that 2 occupants per room, ≥3 occupants per room, and air-conditioner use might decrease the probability of dengue virus infection by 0.43-, 0.45-, and 0.46-fold, respectively.

Our study revealed that people who participated in outdoor sports were at a significantly higher risk of contracting DF than those who did not participate in outdoor sports. This result may be attributed to the preference of residents who participate in outdoor sports to hike and camp in forest margins, tree copses, and natural reserves, which are the original habitats of *A. albopictus* [20]. Therefore, given that participation in outdoor sports increases the risk of exposure to mosquito bites, anti-mosquito measures, such as treating outdoor areas and materials with insecticides, must be adopted.

We also found that human population density was closely associated with dengue transmission. In general, high population density is a risk factor for dengue transmission [21,22]. We found, however, that crowded households comprising ≥2 occupants were at low risk of dengue infection. On the contrary, Velascosalas et al. found that crowded households with more than 1.5 occupants in one room were at risk of dengue infection [23]. We conducted our study in communities wherein individuals lived together in family groups and wherein parents and their young children tended to share one room. Wang et al. reported that 62.48% of Chinese children aged 0–5 years old shared beds with their parents [24]. Our result may be attributed to the following: Parents who share rooms with their children pay additional attention to the use of anti-mosquito measures and the maintenance of good environmental sanitation to protect their children from mosquito bites. Our previous study also showed that married participants had a lower rate of infection than widowed and divorced participants [19]. These results sugges<sup>t</sup> that married groups who reside in one room with ≥2 occupants are at a reduced risk of DF infection. Moreover, high numbers of occupants in one room are associated with the decreased probability of mosquito bites when the number of mosquitoes was fixed.

Shen et al. [25] and Wu et al. [26] reported that yearly average temperatures of more than 18 ◦C would increase the risk of dengue virus infection. Meanwhile, our study indicated that air-conditioner use was a protective factor against dengue infection by reducing the risk of dengue transmission through cooling the indoor environment. In addition, doors and windows are commonly shut during air-conditioner use; this practice could also reduce the chance that mosquitoes could enter the rooms [27].

We found that the poor indoor daylight quality increased the likelihood of infection with DF by 2.27-fold because adult *A. albopictus* prefer to inhabit poorly lit areas over well-lit areas [28,29]. As a result, environments receiving insufficient daylight encourage the density of mosquitoes to increase because they are suitable for the survival of mosquitoes.

Vanwambeke et al. [30] and Kenneson et al. [31] reported that the use of mosquito nets reduced the risk of dengue virus infection. However, similar to Tsuzuki et al. [32] and Loroñopino et al. [33], we failed to find a relationship between the use of mosquito nets and the likelihood of dengue virus infection. The lack of a relationship between this variable and dengue virus infection may be attributed to the following: Mosquito nets are usually used at night. However, *A. albopictus* is active during the day, especially in the early morning and late afternoon [34]. Other studies found that mosquito nets play a protective role in preventing dengue virus infection in rural settings [29]. However, we recruited our study population from urban areas. The good living environment in urban areas, for example, the popularity of air-conditioning and mosquito killing facilities, reduced the demand for mosquito nets. Nevertheless, on the basis of our local experience and some studies' results [30,35], we recommend using mosquito nets not only at night but also during the day [31].

Andersson et al. [36] and Roberto et al. [11] reported that the government's ability and capacity to control the dengue vector has crucial effects on dengue transmission. Community neighborhood committees and property managemen<sup>t</sup> departments in Guangdong Province have organized numerous health remediation activities under the supervision of the relevant health agen<sup>t</sup> or the Centers for Disease Control and Prevention [37] given the high incidence of DF in recent years. These activities have considerably improved the residential living environment and reduced mosquito breeding and may account for the lack of the statistical significance of the variables of domestic sewage disposal, garbage management, and residential surroundings in this study. However, we found that 67.51% of the cases and 60.39% of the controls did not participate in the community hygiene managemen<sup>t</sup> intervention activities organized by neighborhood committees or property managemen<sup>t</sup> organizations. The public health consciousness of the case and control groups must be strengthened because the governmen<sup>t</sup> will be unable to establish a sound prevention system against DF despite having a good macro-control system in place if it lacks the support of the masses.
