I have a longitudinal data set with two waves. Several variables exist in both waves. I want to examine the following question with and OLS regression: "How does perceived family support affect someone's level of depression."

IV: "When I have a serious problem, I can always count of my family for support."(0=Agree; 1=Disagree) --> this question was asked both in the first and second wave. I will use the second-wave variable as the IV and disregard the first wave variable.
DV: Depression Scale (numeric) --> this data was also obtained in both the first and the second wave. I will use the second-wave depression value as the DV and include the W1 depression scale as an IV in my model, as someone's previous level of depression affects their current level of depression.

Beyond that, however, I am not sure which variables (that is, variables from which wave) to include in the model.

Hence:
Family support (W2) + Depression (W1) --> Depression (W2)


Is this actually the best way to design my model/to answer my research question?

Any input would be highly appreciated. Thank you!