Fascinating work, Jin. I would argue (not to be too technical) that even the first level you describe - uncovering an association between food insecurity and smoking - is second generation, as it is getting at why social determinants associated with disadvantage can lead to worse health outcomes. Food insecurity is one aspect of being low SES, and therefore may be (and it sounds like it is) on the pathway from being low SES to increased smoking rates. You are then "peeling back the onion" to get at a greater understanding of how this functions, doing more second generation work, with the goal of an intervention. Food insecurity is an interesting and, as you know, increasingly researched aspect of social determinants, and I mentioned that in the lecture on Tuesday.
I am sure you are aware of the social branding intervention around smoking being done by Pam Ling - I think this draws on many of the same constructs as the barbershop intervention, but takes it to the next level in that it is about social norms, and not connected to the health care system.