I was interested that income level was not considered as part of the potential correlation effect. Perhaps they imply that education level approximates this, b

I was interested that income level was not considered as part of the potential correlation effect. Perhaps they imply that education level approximates this, but we do know that income level is relevant because of the higher "at checkout" cost of actual food. One correlation effect would be the higher income advantage in obtaining and regularly consuming fresh foods and what I suspect would be lower depression rates in general among more affluent participants (this is my postulate, not addressed in the study as far as I can tell). Note: I'm not saying that poor people are all sad and rich people are all happy. My real world experiences run to the contrary in significant ways. I'm just suggesting that amongst these civil servants, there could be an economic feasibility bias that impacts the correlation and would further suggest against causation. EDIT: corrected poor grammar.