A descriptive cross-sectional study was conducted among tribal under five children over a period of 6 months from September 2013 to February 2014 in three randomly chosen tea gardens of Naxalbari block of Darjeeling District. The inhabitants of these tea gardens are exposed to environmental dynamics of poor housing, water supply and sanitation.

Considering the prevalence of childhood morbidity noted in a similar type of study as 41.4% [3], a confidence level of 95%, 10% absolute precision, design effect 2, and 10% non-response rate, the sample size was calculated as 208. However, 208 were rounded off to 210 to get an equal sub-sample size of 70 from each of the gardens. A list of all tribal children in those tea-gardens was prepared with the help of tea-garden officials and local health workers which served as the sampling frame. Systematic random sampling was applied to choose 70 children from each of the tea gardens.

After getting permission from the Institutional ethics committee and tea-garden officials, pilot testing of the questionnaire was done among a small sub-sample of children, to check for consistency.

Data collection was done at the household level using a predesigned and pre tested questionnaire; mothers being the respondents. If any selected eligible study subject was found absent on the day of data collection, the household was revisited up to 2 times. Subjects not available even after those three visits were considered as non-responder.

The mothers were briefed about the purpose of the study and informed consent was obtained from them after assuring confidentiality and anonymity. They were asked about socio-demographic variables (age of child, religion, socioeconomic status, literacy level of mother) and history of any illness during last 15 days for assessment of morbidity. Immunization status was noted from the immunization card.

Every child was subjected to a thorough clinical examination in daylight for assessment of morbidity according to IMNCI guidelines. Body weight was measured with the help of platform weighing machine in case of children >1 year and Salter weighing scale in case of infants which were standardized and calibrated before every measurement. The weight was measured following standard guidelines and measurements were taken to an accuracy of 100 grams. The weight was plotted against their age in WHO growth chart separately for boys and girls and nutritional status was noted.

Collected data was checked for consistency and entered in Microsoft Excel data sheet. Data was analysed by IBM Statistical Package for Social Sciences (SPSS) version 20. It was organized and presented using the principles of descriptive statistics.

Binary logistic regression was applied to test association between morbidity and other epidemiological correlates. In this analysis, morbidity status of the child was used as the dependent variable, where having any morbidity and/or history within last 15 days was taken as 1 and absence of morbidity was taken as 0. The predictor variables used for the analysis include: Age, gender, religion, caste, socioeconomic status, mother’s education, immunization status and weight for age.

The child’s age was categorized into < 1 year and ≥ 1 year; gender as male and female; and the religion were grouped as Hindu and Christian. A relative indicator of household wealth was calculated from the per capita monthly income of the family and categorized as per recommendations of Tendulkar Committee’s recommendation [12] into two classes – Above poverty line (≥ Rs. 673) and below poverty line (< Rs. 673). Immunization status was noted from the card and categorized into completely (up to date) and incompletely immunized according to their age and nutritional status was categorized as Normal and low weight for age (Below 2 SD).

In the present study, 192, under five children were included; the mean age of the children was 31.1 ± 16.4 months and the males (51%) slightly outnumbered the females (49%). Majority of the children belonged to families living below poverty line and following Hinduism. Fourth of the sample studied belonged to lower class and almost 90% of their families were Hindu by religion. 62.5% mothers were illiterate. For 81.6% of the total enrolled children, birth weights were of normal and immunization status was completed for 92.2% of them. But it was seen that only 35.6% of the children were of normal weight for age.

Overall, 38.5% (74 out of 192) of the children had any morbidity, diarrhoea (26%). being the commonest followed by acute respiratory infections (24.5%). Palmer pallor was present in 24.5% children [Table/Fig-1].

Binary logistic regression analysis suggests that the odds of suffering from morbidities were significantly higher among children less than one year of age, female gender, belonging to families following Hinduism, and having malnutrition. Socio-economic conditions, mother’s education and immunization status were not found to be significantly associated with morbidities [Table/Fig-2]. Though children belonging to BPL families and having illiterate mothers had higher odds of suffering from malnutrition than their respective counterparts, the difference was not found to be significant. Surprisingly, immunization status was not found to be significantly associated with morbidities.

Understanding the morbidity and nutritional status of under five children has far-reaching implication for the better development of future generations. In this study, morbid conditions in the form of common illnesses were found among 74 children (38.5%), which was quite less as compared to the studies by Ukey et al., and Dongre et al., showing illness in 41.4% and 72.6% children respectively [3,4]. The reason might be the improving immunization coverage in recent years and also health services and their utilization have increased over past few years.

The leading causes of morbidity among all the children were diarrhoeal diseases (26%), followed by acute respiratory infections (24.5%) and fever (16.7%). Whereas most of the studies conducted in other parts of the country reported ARI as the major cause of morbidity among under five children [3–10]. Such higher proportion of diarrhoea in the study area might be due to lack of supply of safe and wholesome water and also absence of sanitary latrines amongst the vast majority of the houses of the tea garden areas.

Most striking feature was a very high proportion of underweight children (64.4%) in the study area than the findings of other studies in other parts of the country, like a study in Bhubaneshwar by Pattnaik et al., [5]. This may be both due to long-term deprivation as well as recent episodes of diseases.

Results from logistic regression analysis suggest that the odds of morbidities among children less than one year of age, female gender, belonging to families following Hinduism, living below poverty line, illiterate mothers and having malnutrition were higher compared to their counterparts. This pattern could be due to factors such as an increased exposure to contaminated food in these circumstances. A unique feature of the study is that socio economic status and maternal education were not found to be significantly associated with morbidity. This can be explained by the fact that most of the mothers who dwell in the tea gardens are labourers themselves and earn very meagerly. Thus the time spent for rearing children is greatly reduced. The fact that immunization was not associated significantly with morbidities may be attributed to the very low proportion of unimmunized study subjects.

Morbidity was found to be quite high among the tribal children living in tea garden areas. The findings of the present study may have important implications for policy-makers and planners seeking to meet national and international development targets.