Prevalence of water-borne diseases in western India: dependency on the quality of potable water and personal hygiene practices

Unlike urban and semi-urban settlements where the potable water is supplied through a water treatment plant and a distribution network, in rural low-income settings, the provision of the water treatment plant for all villages is not feasible for a developing country like India. The most affordable and reliable way to provide clean drinking water is treatment at the consumer end. This research aimed to assess the occurrence of water-borne diseases based on personal hygiene and the quality of drinking water. Of the households, 4,237 in 15 selected villages were surveyed for personal hygiene using a questionnaire. Water samples were collected from all major water sources in the villages and analyzed for chemical and bacteriological properties. For water and personal hygiene, quality indices were calculated, and a mathematical model was developed using multiple linear regression analysis. The regression concluded that hygiene has a more signi ﬁ cant effect on the occurrence of water-borne diseases than the quality of water. Personal hygiene is one of the health factors neglected by the people speci ﬁ cally in rural India. Therefore, India needs to run campaigns like Swachh Bharat Abhiyaan (Clean India Mission), which was mainly aimed to reduce open defecation, to promote personal hygiene and to reduce the prevalence of water-borne diseases.


INTRODUCTION
The importance of water, sanitation and hygiene (WASH) has been proved to be a crucial factor in the control of water-borne borne diseases ( But JMP, due to the lack of affordable methods to test water quality directly, collects data from national surveys to investigate improvement in water source, which makes the data less reliable (Wang et al. ). Even though JMP conducts various workshops in order to regulate the assessment of indicators by national authorities, which can help in preparing more accurate estimates, but these indicators may lack region-specific desegregation. The government officials and policymakers need to have reliable microscale data with interpretation to develop region-specific plans to improve basic services and tackle disease burden.
In this project, data were collected from Solapur district in the state of Maharashtra located in western India. With low average rainfall ranging from 350 to 680 mm/year, and the study area belongs to one of the dry districts in western India (Yadav et al. ). Ujani Dam constructed on the Bhima River located near Indapur is a major source of freshwater to the Solapur district with an effective storage capacity of 1,517.2 MCM (NRLD ).
The Maharashtra Pollution Control Board approached CSIR-NEERI to assess the prevalence of water-borne diseases due to water pollution in the region. Due to the lack of sources of water and poor economic status, the region is threatened with the prevalence of water-borne diseases like typhoid and cholera. This research aims to understand the dependency of water quality and personal hygiene on the occurrence of water-borne diseases and prepare a model to predict the occurrences of water-borne diseases in the study area. Regression analysis is used to understand the relationship between the variables and to develop a mathematical model (

METHODOLOGY Primary survey and site selection
A preliminary survey was carried out to identify all the rural villages located near the bank of the river Bhima starting from the Ujani Dam to Pandharpur in Solapur district.
Throughout the patch of the river (178 km), the sources of water supply were earmarked, and demographic information of each village was collected. Initially, 26 villages were identified with more than 500 households, and a water source located near the river. Out of these, 14 villages were selected randomly and an additional village which is distant from the river. Since the village distant from the river uses groundwater as a major source of freshwater, it is considered as a control village to evaluate the impact of personal hygiene on the prevalence of water-borne diseases.
After the primary survey, the study was divided into three activities: 1. Identification of sources of drinking water, water quality assessment and calculation of overall water quality index (OWQI).
2. Household survey to study the hygiene practices, occurrence of water-borne diseases and calculation of personal hygiene index (PHI), as well as identification of possible sources of freshwater contamination.
3. Regression analysis to understand the dependency of occurrence of diseases on PHI and OWQI.
Survey of sources of freshwater, water sampling, analysis and calculation of OWQI The selected villages in the study area were remotely located with bad road connectivity; therefore, the preservation and storage of samples before transport to a laboratory was not feasible. Hence, two main parameters were selected, chemical oxygen demand (COD) and fecal coliform (FC), and analyzed at a temporary laboratory setup as per standard methods of analysis (APHA ). During the primary survey, all the major sources of water for the villagers were identified, and samples were collected for analysis.
A method formulated by Singh et al. () was followed for the calculation of OWQI. This method was developed for the Indian subcontinent. COD and FC were classified based on the proposed range. For COD, the range is given as <2, 2-2.9, 3-7 and >7 and for FC, <50, 51-500, 501-5,000 and >5,000. Subindices were calculated using respective index functions provided by Singh et al. ().
where s i is the subindex value for the parameter; w i is the weightage given to each parameter, and w is the total weightage allotted to all parameters. OWQI was constructed by the aggregation of subindices with weightage to COD and FC as 2 and 4, respectively (Singh et al. ).

Wastewater collection and disposal systems
To understand the existing wastewater disposal system in all selected villages in the study area, site visits were conducted.
Wastewater network and disposal system were studied. As most of the villages in the study area were observed to have gutter networks, they were mapped along with disposal sites and the water sources, which are located beside the network, were identified as the potential cross-contamination locations.

Ethics statement
Permissions were acquired from respective local and state administrative government bodies along with the department of health to conduct this survey. The team of surveyors collected data by voluntary enrollment of participants in the study area. To record the response, questionnaire survey sheets, duly signed by respondents after explaining the objectives of study, purpose of data collection, analysis of collected data and privacy rights, were collected. The methodology was approved by the Ethics Board.

Survey for health indicators and development of PHI
The main objective of the survey was to collect information on the hygiene practices and water-borne disease data from district hospitals and village primary health centers. The questionnaire was divided into four main sections. The first section was aimed to obtain data regarding sources of water and treatment methods used in individual households.
The second section was designed to obtain information regarding water handling and storage practices. As main storage practices in the study area are clay utensils and plastic tanks, they are prone to sediment deposition and microbial growth.
Therefore, questions regarding the frequency of cleaning the storage units and the frequency of refilling were included.
The third section was more focused on the defecation practices including the availability of toilets, handwashing techniques and the provision of septic tanks. The last section was dedicated to collect information regarding the reported cases of water-borne diseases for year 2018-19. Diseases include typhoid, cholera, diarrhea and gastroenteritis.
For the development of PHIs of each village, a weightage is given to each of the parameters according to its possible impact on the health, and indices were developed.
where p i is the parameter value for the village, and w i is the weightage to parameter. For the whole village, according to their practices, each parameter was calculated in percentages, and the percentages were used as parameter values for the village.

Regression analysis
To understand the dependency of the prevalence of waterborne diseases on OWQI and PHI and to develop a model, multiple linear regression analysis was used.
where y i is the dependent variable (occurrences of waterborne diseases), x 1 and x 2 are independent variables OWQI and PHI, respectively, β 0 , β 1 and β 2 are coefficients and ε is the error. A design matrix was used to calculate the coefficients. The F-test is used to test if a group of variables is significant jointly, and the following formula was used to calculate the F-value.
where SSR and SSE denote the regression sum of squares and the error sum of squares, respectively; n and k are the number of observations and the number of parameters, respectively. In addition to this, the p-value is also calculated. The F-value tests the significance of a group of variables, whereas the p-value calculates the significance of each parameter separately and it suggests the probability of a value exceeding the F-value (Schaalje & Rencher ).
Therefore, the p-value is calculated to understand the significance of individual variable. For the p-test, the degree of error (α) is assumed to be 5%.

Sources of freshwater and wastewater collection and disposal systems
It was observed in most of the villages, wastewater collection and disposal system were inadequate. The villages with borewells as a main source, shown in Table 1, observed to have a distribution network for the supply of water but without any treatment. Some of the villages with borewell had partial connectivity, and unconnected households   Table 1).

Hygiene practices
Figures 4-7 summarize the hygiene habits in the study area.
In total, 40% of the villagers in the study area use water for domestic purposes including drinking and cooking without any treatment. Among the villagers who treat water before use, 55% use cloth filter, 35% use RO þ UV and the rest 10% use other methods. More than 90% of the villagers refill fresh water daily. In the villages facing water scarcity, it was observed that the villagers tend to store water for a longer period of time and without washing the storage utensils. But the proportion of villagers practicing this technique was found to be very low. Also, 85% of the villagers in the study area use toilets, and 96% use soap for handwashing.
Occurrence of water-borne diseases Figure 8 illustrates reported cases of water-borne diseases, which were observed between 0.25 and 4.18% of the total population with an average of 1.7% in the study area. The most affected age group was kids of age less than 15 years followed by adults of age 30-50 years as shown in Figure 9.
The least affected age group was elderly people more than 50 years of age.

Regression analysis
In this study, PHI and OWQI are considered as independent variables to understand their significance on the occurrence of water-borne diseases, which is the dependent variable.
In Table 2, ANOVA and p-value suggest no significant relationship between OWQI and the occurrence of waterborne diseases. In the case of PHI, ANOVA and p-value signify the relationship with the occurrence of water-borne diseases, and the coefficient of determination is 0.466 which indicates a moderate relationship as illustrated in Table 3.

Uncorrected Proof
To establish a model for the prediction of the occurrence of water-borne diseases based on PHI and OWQI using multiple linear regression analysis, ANOVA indicates that the relationship between a group of variables is significant and reduce the error generated due to less significant variable. Here, the p-value for PHI signifies the importance over OWQI. Using regression analysis in Table 4, the model can be represented as follows: y ¼ 12:33 À 0:318x 1 À 1:147x 2 where y is the occurrence of water-borne diseases, x 1 is the OWQI and x 2 is the PHI.
The line fit plot also indicates the moderate correlation between PHI and the occurrence of diseases, which is a very well-known phenomenon, but Figure 10 illustrates that the source water quality is having less significance on the occurrence of diseases.

DISCUSSION
Results suggest that personal hygiene has significant importance over the quality of water source for the occurrence of water-borne diseases like typhoid and cholera. This study is focused on the quality of source water. The treatment of

CONCLUSION
Even though the relationship between personal hygiene and the occurrence of diseases is obvious, this study emphasizes the importance of it. Infrastructure development alone will have less influence on curbing diseases. A small improvement in personal hygiene can effectively reduce the cases of water-borne diseases. In particular, the study indicates that the household treatment of water combined with good handwashing practices contributed to controlling water-borne diseases. Systematic assessments using the risk-based approach will help identify region-specific health indicators. The intervention campaigns must be developed to address priority risks.