I am working with panel data and use xtlogit command to conduct multivariate regression analysis. For one of my explanatory variables, the data is rather skewed for this dichotomous variable with 57 coded "1" and 7166 coded "0". I am interacting this variable with gender variable, and that result in 35 males and 22 females in one group. I got a statistically significant result for males, but I wonder if it is problematic considering the data being so skewed.

I appreciate your advice.