Q was dichotomized into high and low categories. The cut-off point was one SD below the mean Q, i.e. at the 27.4. While selecting control and farmer we matched for five years age group (15-19, 20-24 and so on). For example, 35 years of farmer might have 39 years of control. That is why; we found statistically significant difference in mean ages
between control and farmers. So, age was adjusted in further analysis. Regression analyses were performed adjusting for body mass index, age and literacy.