Socio-economic and household determinants of malaria in adults aged 45 and above: analysis of longitudinal ageing survey in India, 2017–2018

Background Even though malaria cases have drastically come down in the last decade, malaria remains a serious public health concern in many parts of India. National Framework for Malaria Elimination in India (2016–2030) has been launched with the goal to eliminate malaria by 2030. Understanding the socio-economic and household determinants of malaria at the national level will greatly aid India’s malaria elimination efforts. Methods The data from Longitudinal Ageing Survey of India (LASI) Wave 1 (2017–2018) survey comprising 70,671 respondents ≥ 45 years across all the States and Union Territories were used for the analysis. Simple and multiple logistic regressions were used to obtain the unadjusted and adjusted odds ratio respectively of the socio-economic and household variables. Results The major socio-economic variables that increase the likelihood of malaria are caste (‘scheduled tribes’), low education levels and rural residence. The scheduled tribes have 1.8 times higher odds of malaria than the scheduled castes (AOR: 1.8; 95% CI: 1.5–2.1). Respondents with high school education (6–12 grade) (AOR: 0.7; 95% CI: 0.6–0.8) and college education (AOR: 0.5; 95% CI: 0.4–0.6) had a very low risk of malaria than those with no school years. Rural residence and occupation (agriculture and allied jobs) also increases the odds of malaria. The major housing determinants are household size (≥ 6), housing type (kutcha), use of unclean fuel, outside water source, improper sanitation (toilet facilities) and damp wall/ceiling. Conclusions The study has identified the major socio-economic and housing factors associated with malaria in adults aged 45 and above. In addition to vector and parasite control strategies in the tribal dominated regions of India, improving literacy and housing conditions may help India’s malaria elimination efforts.


Abstract Background
Even though malaria cases have drastically come down in the last decade, malaria remains a serious public health concern in many parts of India. National Framework for Malaria Elimination in India (2016-30) has been launched with the goal to eliminate malaria by 2030. Understanding the socio-economic and household determinants of malaria at the national level will greatly aid in India's malaria elimination efforts.

Methods
The data from Longitudinal Ageing Survey of India (LASI) Wave 1 (2017-2019) survey comprising 70671 respondents ≥45 years across all the States and Union Territories were used for analysis.Simple and multiple logistic regressions were used to obtain the unadjusted and adjusted odds ratio respectively of the socio-economic and household variables.

Results
The major socio-economic variables that increase the likelihood of malaria are caste (Schedule Tribes), rural residence and low education levels.The scheduled tribes have 1.8 times higher odds of malaria than the scheduled castes (AOR: 1.8; 95% CI: 1.5-2.1). Respondents withhigh school education (6-12 grade) (AOR: 0.7; 95% CI: 0.6-0.8) and college education (AOR: 0.5; 95% CI: 0.4-0.6) had a very low risk of malaria than those with no school years. Rural residence and occupation (agriculture and allied jobs) also increases the odds of malaria. The major housing determinants are household size (≥6), housing type (kutcha), use of unclean fuel, outside water source, improper sanitation (toilet facilities) and damp wall/ceiling.

Conclusions
In addition to vector and parasite control strategies, improving literacy and housing conditions, especially in tribal dominated regions will greatly aid in India's malaria elimination efforts.

Background
Globally, vector borne diseases (VBD) account for > 17% of all the infectious diseases, and account for > 700,000 deaths annually [1]. Malaria is a major vector borne disease, and is a serious public health concern in many parts of India.Malaria situation in India is complex with varied distribution of P. vivax (Pv) and P. falciparum (Pf) [2]. As per WHO, 93% of the population in India are at risk of malaria [3]. From 2000 to 2017, malaria morbidity and mortality in India have declined by 59% and 89% respectively [4]. In India, malaria is transmitted by several Anopheles spp., and the geography determines the primary Anopheles vector; An. stephensi, An. culicifacies, An. dirus, An. uviatilis, An. minimus and An. sundaicus are considered to be the primary malaria vectors in India [5]. Eliminating malaria throughout India by 2030 and 2. Maintaining malaria-free status in regions where malaria transmission is disrupted, and preventing reintroduction of malaria [7]. Furthermore, Malaria Elimination Research Alliance has been launched under the umbrella of Indian Council of Medical Research (ICMR) to "identify, articulate, prioritize and respond to the research needs of the country to eliminate malaria from India by 2030" [4].
India's malaria control strategies focuses on effective vector control using indoor residual spraying, long lasting insecticidal nets, chemical insecticides, bacterial pesticides and larvivorous sh [8]. Chemotherapy using artemisinin based combination therapy for Pf and chloroquine/primaquine for Pv is the major treatment strategy under the national drug policy [9].
Malaria transmission is in uenced by several factors, including socio-economic and demographic characteristics of the study area [10]. In addition, housing factors have also been shown to play an important role in malaria transmission [11,12]. In India, studies on socio-economic and household determinants of malaria are very limited, and are focused on selected districts [10,[13][14][15]. Type of house, toilet facility and water source were the major housing risk factors [13], while social groups, family size [10], monthly income [14] were some of the key socio-economic determinants reported in these Indian studies.
A pan-India study on the socio-economic and household determinants of malaria will give important insight on the major risk factors, and greatly aid in the national malaria control and elimination efforts. Longitudinal Ageing survey of India-1 (LASI-1) carried out across India from 2017-2018 provides important insight into various health parameters of elderly individuals (> 45 years), including the socio-economic and housing conditions of self-reported cases of malaria in the past 2 years before the survey. This LASI data set from 70671 individuals was used to analyze the key socio-economic and housing determinants of malaria, and the results are detailed.

Outcome variable
The outcome of interest is malaria, and was based on the following question: 1.In the past two years, have you had malaria? The options were: 1. Yes, and 2.
No. The response was coded as a binary variable (No-0: absence of malaria; Yes-1: presence of malaria).

Socio-economic status and demographic variables
The socio-economic status (SES) and demographic variables used for this analysis are age-group (45-54 years, 55-69 years and 70+ years), sex (male/female), place of residence (rural or urban), income (poorest, poor, middle, richer and richest), educational level (0 school years, 1-5 school years, 6-12 school years and college), caste (SC, ST, OBC and forward caste) and work (not working, agricultural and allied, self-employed and wage/salary worker).

Statistical Analysis
A frequency distribution table was prepared for all the variables used in this study. Prevalence of malaria with each of the SES (age group, sex, residence area, education level, MPCE quintiles, caste, and work) and household variables (household size, type of house, location of water source, toilet type, cooking fuel, and damp wall or ceiling) were reported. Simple logistic regression was used to obtain the Unadjusted Odds Ratio (UOR), and those independent variables found signi cant (P<0.05) were included in the multiple logistic regression to arrive at the Adjusted Odds Ratio (AOR) of SES and household variables affecting malaria. Weights were applied during analyses. All analyses were performed using STATA MP statistical software version 16.

Spatial Analysis
The malaria prevalence (%) data was added as attribute in State and Union Territories (UT) boundaries of India using ArcGIS software. The histogram of the malaria prevalence data was analyzed, and was used to determine the four class intervals for mapping; the national average of malaria prevalence is 7.91%.

Prevalence of malaria in adults ≥45 years
The frequency distribution of all the study variables is shown in Table 1. The prevalence of malaria in adults ≥ 45 years is 7.9%. The self reported prevalence of malaria across all States and UT of India is shown in Fig. 1. Malaria prevalence is high in Central and Western India when compared to the South, North and Eastern regions. The states with very high (> 15%) prevalence include Rajasthan (23.3%), Chhattisgarh (21%), Madhya Pradesh (20.9%), Gujarat (16.4%) and Jharkhand (16.3%). The above national average (10-15%) prevalence was reported in the States/UT surrounding Central India─Dadra and Nagar Haveli (14.3%), Haryana (13.3%), Odisha (11.6%), Uttar Pradesh (10.6%) and Bihar (9%). The States/UT to exceed 5% prevalence in South are Andhra Pradesh (5.5%) and Andaman & Nicobar Islands (6%), and in the North-East, it is Arunachal Pradesh (10.7%) and Meghalaya (5.8%). The northern-most States/UT (Jammu & Kashmir, Himachal Pradesh, Punjab and Uttarakhand) show very low (0-5%) prevalence of malaria. The prevalence of malaria associated with socio-economic status (SES) variables and household conditions in adults ≥ 45 is shown in Table 2. Malaria prevalence was similar in all age groups, and was slightly higher in males (8.1%) than in females (7.8%). Malaria was higher in the rural (9.6%) than in urban (4%) respondents. The prevalence of malaria was highest in the ST population (15%) when compared to the other social groups. The prevalence of malaria reduced with increasing levels of education (9.7% in illiterates vs. 3% in the college educated) and income of households (9.2% in poorest vs. 6.9% in the richest). Malaria prevalence was highest in those working in agriculture and allied jobs (10.5%), and was lowest in wage/salaried workers (5.2%). Malaria was highest in the households where there were ≥ six members (9.5%), and in those living in kutcha houses (12%). The prevalence of malaria was more in adults with no source of drinking water within the dwelling (10.6%) than those residents who had water source within their residence (6.9%). The prevalence of malaria was more in households (11.7%) with poor sanitation (unimproved toilet facility) compared to households (5%) that had improved sanitation facilities. Malaria in households using unclean fuel was 11.2%, and in households with damp wall/ceiling, it was 9.6%.   [19]. On the other hand, Mizoram is considered to be one of the highly malaria endemic states in India [20,21]. One explanation for the unexpected numbers could be the study site where the survey was undertaken. For example, in Mizoram, the district of Aizawl reported 57 malaria cases in 2018, while the malaria-endemic districts of Lawngtlai, Lunglei and Mamit reported 2222, 1092 and 772 respectively. Therefore, if the survey was carried out at Aizawl, the capital of Mizoram, the self reported malaria cases will be lower. Another possibility could be the adults in Mizoram may be asymptomatic due to various types of adaptive or acquired immunity [22]. In sub-Saharan Africa, many adults who harbour the parasites rarely show clinical symptoms [22]. At Mamit, the average annual parasite index (API) from 2010 to 2018 was 34.4 (34 cases / 1000), one of the highest in the country [20]. During 2014 to 2015, there was a big spike in malaria cases in Mizoram, and in Mamit district, nearly 50% of the total population (8766 cases out of 17731) were affected in 2015 [20].
Despite the signi cant strides India has made in decreasing malaria mortality and morbidity in the last two decades (from 2000 to 2019, malaria cases and deaths have declined by 71.8% and 73.9% respectively) [23], malaria remains a serious public health issue in several parts of the country. The NVBDCP has developed a comprehensive strategic plan to achieve malaria-free India by 2030 [24]. For devising effective malaria control and elimination strategies, understanding the socio-economic and household variables that affect malaria transmission is imperative. Our analysis indicates rural residence, occupation (agricultural and allied), education levels (illiterates and primary), caste (ST), household size (≥ 6), sanitation (poor toilet facility), unclean cooking fuel, water source not in dwelling, damp wall/ceiling are the major socio-economic and household risk factors that affect malaria transmission.
Not surprisingly, ST population are at a higher malaria risk. Jharkhand, Chhattisgarh, Madhya Pradesh, Odisha and malaria-endemic NE States (Tripura, Meghalaya, Mizoram and Arunachal Pradesh) have a high ST population. The geographical terrain that includes many forested areas, poor accessibility, frequent natural hazards, perennial P. falciparum transmission, very e cient anthropophillic vectors, and socio-cultural practices greatly hinder malaria control efforts in many rural tribal areas of India [2,25]. The orthodox health beliefs of tribal population have restrained them from accessing health services despite them being highly vulnerable to various health hazards, including malaria [25,26]. In addition, the dense forest cover and high rainfall in the tribal belts are conducive for mosquito breeding, and malaria transmission [25,26]. Many of the tribal pockets where malaria is endemic are characterized by poor housing conditions. Residents in kutcha houses have higher odds of malaria; kutcha houses may have holes and gaps that may allow easy entry of mosquitoes. This is in line with earlier Indian studies where kutcha houses/walls made of dung and earth have shown to be a risk factor for malaria [10,13]. Positive association between mosquito bites / day and bamboo houses has been reported in Assam, India [14].
Literacy has a negative association with malaria; illiterates and those with just primary education have higher risk of malaria. This was expected as literacy gives a better understanding of infectious diseases and the protective measures required. However, earlier studies [10,14] in India did not nd an association between education and malaria risk. In Yadav et al, 2014 [14], the sample size was just 71 households, while in Sharma et al, 2015, only no schooling, primary and secondary grades were included, and college education was not included in the education characteristics. ST who are at higher risk of malaria have lower literacy rate (59%), when compared to the national average of 73% [27].
Malaria risk was higher in those who carry out agricultural and allied activities when compared to respondents who are self employed or get wage/salary. Agricultural activities require signi cant time to be spent outdoor, and consequently are highly exposed to malaria bites. A recent study from Mandla district in Madhya Pradesh shows households having own farmlands to have a signi cant association with malaria [13].
Size of the household (> 6 members) contributed a signi cant risk on the malaria incidence. Family size/number of people in the house/number of people per room have been shown to be an important risk factor for malaria [10,13,[28][29][30][31], as crowding attracts more mosquitoes due to strong olfactory signals [32].
As observed in studies carried out in India [10,13,33], Indonesia [34], and sub-Saharan Africa [35], access to outside water source is a major risk factor of malaria, as dependence on outside source for water, especially in dusk and dawn increases the chances of mosquito bite. Furthermore, households using tubewells as outside water source have risk of malaria [13], as tube-wells are suggested to have more stagnant water around them due to improperly maintained drainage facilities [36]. Poor toilet facility (sanitation) is another important household risk factor of malaria, and is in-line with earlier studies carried out in India [13] and elsewhere [33][34][35]37]. Use of unclean cooking fuel is also a major risk factor (odds increase by 1.5 times) for malaria. In addition, damp wall/ceiling was also associated with increased malaria risk; damp walls favour indoor resting of mosquitoes [38]. The three household determinants: outsidewater source, improper toilet facility and unclean cooking fuel increases the likelihood of bitten by mosquitoes outside the house. Increased time required for outdoor cooking using unclean cooking fuel could be a reason for its higher odds. Free clean cooking fuel (liquid petroleum gas connection) has been given to > 80 million Indian households through the Pradhan Mantri Ujjwala Yojana (PMUY) scheme [39]. Through this scheme, it is expected that 80% of the households will have clean cooking fuel by 2019 [40]. In addition to improving the standard of living, the PMUY scheme may also help in malaria control and elimination efforts.
Malaria is considered to be a disease of the poor [41][42][43], and several studies have shown signi cant association between poverty and malaria [13,[44][45][46]. There are also studies that have shown no association between malaria and socio economic status of the household [13,44,47,48]. Interestingly, even though prevalence of malaria is higher in poorest, after adjusting the other socio-economic variables, richest were found to have slightly higher risk of malaria than the other economic categories. Urban malaria is predominantly caused by Pv, and as this is a pan-India study, a higher proportion of respondents positive for Pv could have been from urban cities, and are likely to be socio-economically forward. For example, Uttar Pradesh, the most populous State in India has predominantly Pv [19]. Furthermore, richest, especially in urban cities may get tested promptly, and report accurately. However, the socio-economic-housing risk factors like rural residence, caste (ST), education levels, housing conditions, sanitation, unclean cooking fuel, improper water source and damp wall/ceiling strongly suggest poverty to be a risk factor for P. falciparum malaria, especially in the tribal dominated States of Jharkhand, Chhattisgarh, Madhya Pradesh, Odisha, Mizoram, Tripura and Meghalaya.
The major limitation of the study is that it is limited to adults ≥ 45 years old. Malaria affects all age groups, and this study captures only a particular age group. Furthermore, as malaria prevalence is self-reported, the accuracy cannot be veri ed.
Overall, the study gives important insight on socio-economic and housing determinants of malaria. In parallel to parasite and vector control strategies, improving the socio-economic and living conditions, especially in malaria dominated tribal pockets would greatly accelerate the malaria elimination efforts.