As it is shown above, the variable
agegphas 6 groups (i.e., 25-34, 35-44) which has different alcohol intake and smoking use combinations. I think it would be interesting to transform this dataset from long to wide and to create a column for each age group and show the respective cases. Let see how the dataset will look like.
dt %>% spread(agegp, ncases) %>% slice(1:5)
Click through for a few additional transformations.