Dear users,

I am using National Longitudinal Survey of Youths (NLSY). My research question is:

Does formula fed in infancy increase the likelihood of obesity in the late childhood?

In my dataset, I have two groups of children who are formula fed (Yes) and who are not (NO). A binary variable in this case.

I would like to attribute the over time variation in obesity to formula fed using the following:

Obesity_it=b_0+b_1 (formula feed)_i + b_2 (Controls)_it+ Error_it

Now, the variable formula feed is time-invariant and happened once in time (only infancy) and the fixed effects model cancel out that time-invariant variable. What should I do to estimate explicitly the effects of formula feed on obesity?

Your timely and illustrative response is much appreciated.

Daryosh