Statistical Analysis Plan - help!

#1
I'm doing a summer research project on trends of anxiety and depression. Unfortunately, my preceptor has disappeared in the meantime and I'm left to fend for myself. I've done an extensive literature review, and now have to develop a Statistical Analysis Plan (SAP). While I understand that there are loose guidelines for this, I don't have much experience in statistics, biostatistics or epidemiology. Maybe you guys can help me, I'd really appreciate your input!

The dataset is comprised of 3 surveys in 2000, 2005, and 2010 and symptoms of mental health issues are included, and using psychometric tools we can determine if an individual is suffering from anxiety or depression. I want to prove that prevalence rates of anxiety and depression are significantly decreasing after a war, in a sample of about 1800 people (in total, across the 3 surveys), and stratify this information across sociodemographic characteristics (age, gender, income, education). Here is where I get confused: Men and women have pretty different rates due to behavioral differences. In addition, anxiety and depression are commonly comorbid. How would I set up the models? Should I have two separate statistical analysis plans (one for anxiety, one for depression), with each having two models (one for women, one for men)? From there, what would be a proper test; it seems like chi-square and multiple linear regression would be proper tests.

In addition, if anyone else has ever created a SAP, please give me some advice on what to include. Thanks in advance for your help!