At the same time, one could probably get an ob/gyn, a psychiatrist and a statistician together for a retrospective study - even if it’s a simple questionaire. … It could simply analyze feelings, thoughts etc. … Clinical depression is evaluable, as is anxiety, PTSD, and one can measure drug intake if your state/province has a DPIN system. Other out-comes too might be evaluated in terms of 3, 5, and 10 year morbidity, mortality etc. (i do believe some have been done with some equivocal data).
True, with the first. Though the problem with questionaires is that you have to be extremely careful with the wording. I fear that any study on this can very easily be manipulated by the according interest group who launched it. So, you could make a study testing depression etc., but that then must be done to women who have given birth, have aborted, have not been pregnant, and no question should somehow relate to pregancy.
Then, after evaluated for these three different groups, you could start to make some statitics out of that.
But, i fear that the above-mentioned statistics did not pay enough attention to that, something i conclude from the “30% to 60%”, an error i can not accept when having 1000, say “datapoints”.