Risk factors for bat contact and consumption behaviors in Thailand; a quantitative study

An analytical cross-sectional study was conducted to identify the factors associated with bat contact and bat consumption behaviors among persons who lived in or near areas of high bat density within at least 6 months before the study. The questionnaire interviews were conducted from May 2016 to July 2017. The Chulalongkorn University Institution Review Boards (IRBs) and Chiang Rai provincial health office approved the human study (Ref No. 034/59 and 26/ 2559). This study obtained agreements from local administrative offices and the chiefs of villages to conduct data collection at the study sites. Written informed consent forms were signed by the participants after they received information about the objectives of the study and prior to the interviews.

Different sampling methods were applied to select the study sites and study populations. The whole country was first stratified into four regions; central, northern, northeastern and southern Thailand. Provinces, districts and villages in each region were then purposively selected based on; 1) a high density of bats in the villages/study areas, 2) a potential bat-human interface was observed by researchers during scoping visits, discussed with some villagers and/or local authorities, and 3) information obtained from the relevant local and national authorities such as the Ministry of Natural Resources and Environment (MNRE), and the Ministry of Public Health (MoPH). Lastly, a simple random sampling method (SRS) was used to select respondents from official household registry records that were obtained from the local health promotional hospitals in the villages.

In this study, 10 provinces were selected as representatives of regions in Thailand that have different living characteristics and practices. The provinces were Ang Thong, Ayutthaya, Lopburi and Saraburi (central provinces); Chiang Mai and Chiang Rai (northern provinces); KhonKaen and UbonRatchathani (northeastern provinces); and Krabi and Surat Thani (southern provinces) (Fig. 1 and Table 1). Bats from each study site were collected for species identification using physical characteristics and DNA sequence variations in mitochondrial cytochrome-b (CytB) [21].

The inclusion criteria for respondents were males or females between 20 and 75 years of age who had lived in the selected areas for at least 6 months before data collection and were willing to participate in this study. The sample size calculation for this study was as follows:

Here, p = proportion of participants with bat contact. Based on a previous study, 23.00% of respondents reported bat consumption at some time in their lives [22]. The variable p is 0.23 for this calculation, z = 1.96 (95% confidence interval), and d (margin of error) = 5%. Therefore, the calculated sample size was 340. To increase the power, the sample size was elevated to 626.

Simple random sampling (SRS) was used on the lists of the respondents retrieved from health promotional hospitals in the study areas. After SRS of respondents was conducted, the participants were contacted by trained researchers for data collection. Using a standardized questionnaire, a face-to-face interview was conducted with the respondents in a place that was not too secluded but still free from disturbances.

A questionnaire was designed to collect information on bat contact and bat consumption behaviors. This questionnaire was modified from previous reports [10, 23, 24]. The questionnaire interview was administered to collect information on sociodemographic factors, knowledge, attitudes, practices and perceptions for bat contact and bat consumption behaviors. The questionnaire was pretested with 30 respondents in a district with the same environmental and population characteristics as the actual selected sites. The questionnaire was refined per the pretest results. Field researchers were trained in its administration and in conducting questionnaire interviews with written informed consent forms obtained prior to the interviews. After the interviews, the information was checked for validity and precision before it was entered into SPSS Software Version 22 (Chicago, IL).

Independent variables (n = 23) were included in the analysis. The variables were based on previous relevant studies [10, 25‐29]. The independent variables, including the sociodemographic information, knowledge, attitudes, practices, and perceptions were as follows:

The dependent variable was a report of either one of bat contact experiences by the study respondents. List of bat contact experiences included the following:

The interviewed data were entered into and analyzed by SPSS software version 22 (Chicago, IL). After data cleaning, the dependent and independent variables were assessed. Data were analyzed in 2 steps. First, bivariate analysis was performed, in which the degree of association of each variable was computed and each of the independent variables was analyzed separately. Second, a stepwise logistic regression model was constructed including independent variables that had p values ≤0.15 results in bivariate analysis. This second step used p ≤ 0.05 as the cutoff point for identifying statistically significant variables.

In this study, 626 respondents participated in the questionnaire interview. These respondents from 37 villages of 16 districts in 10 provinces in Thailand (Fig. 1 and Table 1). Bats species from each study site were identified: Chiang Mai (Scotophilus heathii and Megaderma spasma), Chiangxs Rai (Taphozous melanopogon), Lopburi (Taphozous melanopogon and Chaerephon plicatus), Khon Kaen (Chaerephon plicatus), and Ayutthaya, Ang Thong, Saraburi (Pteropus lylei). Of the 626 respondents who participated in the questionnaire interview, 228 (36.42%) respondents from 10 provinces reported having experienced bat contact during the past 6 months before data collection. Moreover, 96 (15.34%) respondents in 9 out of 10 provinces reported having consuming bat meat in the past 6 months.

The 626 respondents were 285 females and 341 males. The mean age of the respondents was 47.58 years. Approximately 21.09% of respondents were aged > 36 years. Most of them lived in rural settings (93.29%) and were married or cohabiting (84.03%). Respondents worked as farmers (rice, grains or vegetables) (35.78%), followed by temporary employees (22.68%), government officers (10.54%), housewives (7.03%), and shop vendors/owners (5.59%). Most of the respondents (65.65%) had attained educational levels lower than secondary school, and 80.35% had families composed of more than 2 persons. Most respondents had a monthly family income of ≤15,000 Baht (500 USD) (69.97%). Most respondents owned a motorcycle (91.69%), while fewer owned a car (50.16%).

In this study, 228 respondents in 10 provinces reported a bat contact experience during the past 6 months before data collection. Among those 228 respondents, 56.14% reported encountering live bats in a house, the community or a tourist location, while 42.10% reported eating bats for food, 34.21% found bat guano in a house or the community, 28.51% found dead bats in the house, 20.18% participated in bat guano mining/collecting, 19.74% cleaned bat carcasses from a house or the community, 10.96% used bat guano as fertilizer, 17.98% were involved in other contact activities (e.g., took bats from nets), and 7.02% been bitten by a bat in the past 6 months (Table 2).

In this study, the questionnaire interview contained 10 sociodemographic questions and 13 questions related to the knowledge, attitudes, practices and perceptions regarding bat contact experiences. Bivariate analysis was performed to determine the associations between bat contact experiences and twenty-three independent variables. Of the 23 independent variables, 11 variables had significantly associations in the bivariate analysis, using p ≤ 0.15 as a cutoff point: sex (male), age group (> 36 years), occupation (agriculture-related occupation), educational attainment (< secondary school), family monthly income (< 15,000 Baht), no concerns about getting diseases from bats, considered bats to be economically beneficial to the community, believed it is safe to eat bats, allowed children to play with bats, and felt safe in areas where bats live (Table 3). All 11 variables were included in the stepwise logistic regression analysis for bat contact experiences. The results showed that sex (male) (OR = 1.56, 95% CI 1.09–2.28, p = 0.014), educational attainment (< secondary school) (OR = 1.45, 95% CI 1.02–2.18, p = 0.041), and considered bats to be economically beneficial to the community (OR = 3.18, 95% CI 2.03–4.97, p < 0.001) were statistically significant associated with bat contact experiences, while occupation (agriculture-related occupation) (OR = 0.54, 95% CI 0.37–0.79, p = 0.002), believed it is safe to eat bats (OR = 0.58, 95% CI 0.37–0.93, p = 0.023), allowed children to play with bats (OR = 0.65, 95% CI 0.44–0.96, p = 0.031), and felt safe in areas where bats live (OR = 0.56, 95% CI 0.38–0.86, p = 0.007) were statistically significant protective factors against bat contact experiences (Table 4).

Among the 626 respondents, 96 respondents (15.34%; 28 females and 68 males) reported eating bats in the past 6 months. Of the respondents who reported contacts with bats, 42.10% (96/228) reported eating bats in the past 6 months. They also reported that bats were hunted, butchered, slaughtered, and cooked by several categories of persons (e.g., husbands, wives, neighbors, children, or hunters) in their communities. In this study, 170 out of 626 respondents (27.16%) reported having eaten bats in their lifetime, of whom 62.35% (106/170) reported consuming bats more than 10 years ago, 11.18% (19/170) who reported eating bats during the last 1–10 years, 20.00% (34/170) who reported eating bats within the past 12 months, 2.94% (5/ 170) who reported eating bats in the past month, 2.35% (4/170) who reported eating bats within week of data collection, and 1.18% (2/170) who reported eating bats within the preceding week.

Of the 96 respondents who reported eating bats, 70.83% were male and 83.33% were < 36 years old (mean age of respondents 47.21 years old). Most respondents were married or cohabiting (85.42%) and had more than one child (69.79%). The occupations of respondents were farmers (rice, grains or vegetables) (39.58%), followed by temporary employees (29.17%). Most respondents had attained an educational level < secondary school (79.17%) and had families composed of more than 2 persons (76.04%). Most of the respondents (66.67%) had monthly family income < 15,000 Baht (500 USD). Some respondents owned a car (42.71%), majority owned a motorcycle (96.88%). Of the 96 respondents, those in the northern region, Chiang Mai (28.13%) and Chiang Rai (13.54%), reported more bat consumption than those in other provinces, including those in the northeastern region, Ubon Ratchathani (18.75%), southern region, Surat Thani (16.67%) and Krabi (14.58%). On the other hand, respondents in the central region reported less bat consumption (Table 1).

With regard to the details of bat consumption behaviors, we found that 26.04% (25/96) of respondents reported killing bats themselves, followed by vendors (15.63%), hunters (13.54%), and neighbors (11.56%). The participants reported that no children aged 0–10 years old killed bats. With respect to preparing and cooking bats for food, 34.38% (33/96) of respondents prepared bat meat by themselves, followed by spouses (14.58%), neighbors (11.56%), and children (1.04%). Persons reported that the bats were eaten by spouses (48.96%), neighbors (39.58%), and children (13.54%). Most respondents reported eating cooked bat meat (98.96%, 95/96), while only one respondent, a 63-year-old female from Chiang Rai province, reported eating raw bat meat. Bats were obtained by hunting in caves (30.21%), purchasing from local markets (19.79%), hunting by themselves (13.54%), and hunting by neighbors (8.33%). Dishes containing bat were “Kang Om” (spicy vegetable soup) (18.75%), followed by spicy stir fry (13.54%), “Kua Kling” (dry spicy fry with herbs), and deep-fried bat meat (8.33%). The preferred recipes varied among provinces or regions; for example, people in the northern region, Chiang Mai, preferred to cook “Kang Om” while those in southern provinces, Krabi, preferred spicy stir fry to other dishes. The field investigation showed that a typical bat preparation process included butchering the bats, selecting the parts for consumption, adding additional herbs for flavor, and boiling the ingredients in a large pot (Fig. 2).

In this study, 23 variables (sociodemographic factors, knowledge, attitudes, practices and perceptions) were involved in the bivariate analysis, and eleven independent variables were eligible for the stepwise logistic regression analysis. These included sociodemographic variables (sex, educational attainment, owning a car, and owning a motorcycle) and variables pertaining to knowledge, attitudes, practices and perceptions (knowledge of one can contract diseases by drinking water from the places where bats live, knowledge of whether it is safe to eat bats, knowledge of whether there are laws pertaining to hunting bats for consumption, practice of allowing their children to play with bats, and attitude of feeling safe in areas where bats live) (Table 5). In the stepwise multiple logistic regression analysis, 5 independent variables were statistically significant (p ≤ 0.5). Sex (male) (OR = 2.48, 95% CI 1.49–4.11, p < 0.001) and educational attainment (≤ secondary school) (OR = 2.21, 95% CI 1.27–3.85, p = 0.005) were statistically significant related to bat consumption behaviors, while knowledge of whether bats are safe to eat (OR = 0.04, 95% CI 0.01–0.25, p = 0.001), knowledge of whether there are laws related to hunting bats for consumption (OR = 0.35, 95% CI 0.18–0.71, p = 0.004), and allowing children to play with bats (OR = 0.51, 95% CI 0.31–0.81, p = 0.006) were statistically significant protective factors against bat consumption behaviors (Table 6).