Minimizing Bias and Confounding

The mayor of your city wants to allocate funds to open several curbside healthy food markets/ farmers markets based on the hypothesis that people who utilize curbside healthy food markets/farmers markets regularly (defined as once a week) for their groceries have healthier weight and are likely to control of their weight as compared to people who do not utilize them regularly.

As an epidemiologist you request and succeeded in getting a small amount of funds to conduct an epidemiological study to test the above hypothesis. How would you design this study? What are some of the potential confounders and or biases that can influence the results of study? How can you minimize them and report the results objectively?