Of course, it is not a surprise at all to find that journalists are almost all incapable of reading scientific papers and then reporting their contents sensibly. For a start, your average journalist thinks everything has to be "shock! horror! doom!" so that we will read it, and they will get an award for their journalism. Secondly, they don't have the education that the scientists writing the paper have, and therefore can't really understand it.

So it was with yesterdays's big shock horror "Antidepressants don't do anything!" story.

The first I heard of it was my wife complaining about a story she had just heard on the radio, yesterday morning. She had realised, being rather more clever than your Walrus, that what the story should have actually said was "Antidepressants only work on people who are actually depressed". I thought I would look into it, but after a couple of articles on Google News I thought I would try to find the actual paper.

Here it is...

Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration

Notice that it is a Meta-Analysis. That is to say, a review of existing research, to see if there is a trend emerging from all the research on a subject. So it isn't exactly news, or actually new.

Notice also, that it is only concerned with data submitted to the FDA, which is in America, and not actually the only place research can be submitted to. If there has been a meta-analysis of data submitted in Europe to European authorities, we can not find out about it from the article under consideration.

Anyway, I read the paper. As my wife had surmised from the radio story, it does indeed actually say that, yes, antidepressants do work for people who are seriously depressed. Well, actually, we happened to know that was true. And it said that in tests, it was apparent that the "placebo effect" occured, so that some depressed people who were given fake drugs sometimes showed an improvement.

Now you can not jump from there, as all the non-scientific journalists have, to saying that therefore antidepressants are of no use. Apart from anything else, even depressed people sometimes stop being depressed without any medication, and that is going to have an effect in the tests.

For a proper report on what the analysis actually shows, you can try a story by journalists with some scientific understanding.

I think that what the press may be hinting at is the well know fact that many, many Americans, and others, are taking Prozac and similar drugs to feel better, without actually having severe depression at all. I think the hideous word "lifestyle" becomes involved here. And if anyone is likely to demonstrate the "placebo effect", it is likely to be somebody who takes a drug they don't need because their friends have told them it will make them feel good.

And finally, as they say on the news... A sensible report on MSNBC (whatever that is) confirms the truth of the matter - Think twice before you dump antidepressants! If you actually need them for real depression, continue to take them.

And if you are a shock horror reporter and don't like my tone, well, see if I care, fool.