There's a pretty significant difference between a depressed state and Major Depressive Disorder. My understanding is that people get depressed from all kinds of life events, but that they are typically resilient enough to get better on their own. In contrast, Major Depressive Disorder is the black dog that's always there waiting to pull you down.

You probably don't need a pill if you're depressed because your husband died. It's normal to be depressed, and what you probably need in that case is some kind of social structure or coping mechanism that doesn't involve ruminating about how much you miss your husband.

It's not a disorder to feel sad sometimes.

That's not a long-term study. It's a one-off 'meta-analysis'.

Yes, it's not a long-term study -- it's a summary of 522 studies involving a total of 116,477 people treated for 8 weeks on average.

The authors do note that 8 weeks is not long enough to see some antidepressant side effects, which is a problem, but if they see evidence of improvement after 8 weeks, that's quite different from having no evidence of improvement at all.

Unless you mean to suggest that over the very long term, say several years, people on antidepressants end up no better off? I would be interested to know if that has been studied.

I don't know if they are better off or not, but several years, at least, is what I thought of when I read the original poster refer to "long term".

I do not consider 8 weeks to be "long term".

> if they see evidence of improvement after 8 weeks, that's quite different from having no evidence of improvement at all.

There may show that there are short-term improvements, but it says nothing about the medium to long-term effects on depression.

It's therefore worthless for being able to perceive any of the longer-term negative side effectives of the drugs.

It almost seems like a study that the drug companies would find as favourable advertizing for their drugs. It makes me suspicious that it may have been funded by drug company money.

> It almost seems like a study that the drug companies would find as favourable advertizing for their drugs. It makes me suspicious that it may have been funded by drug company money.

The funding is openly disclosed:

> This study was funded by the National Institute for Health Research (NIHR) Oxford Health Biomedical Research Centre (BRC-1215-20005) and the Japan Society for the Promotion of Science (17K19808). The views expressed are those of the authors and not necessarily those of the UK National Health Service, the NIHR, or the UK Department of Health.

One of the authors, John Ioannidis, is famous as a crusader for better evidence and better statistics in medicine; he's the author of the famous "Why Most Published Research Findings Are False" paper about the potential for false positives for medicine. I would be willing to trust the statistical evidence, though I think you're right to be concerned about 8 weeks as too short a time period.

> It makes me suspicious that it may have been funded by drug company money.

That’s the problem. Always has been. Always will be.

Quoting from your link: "We recorded the outcomes as close to 8 weeks as possible for all analyses. If information at 8 weeks was not available, we used data ranging between 4 and 12 weeks (we gave preference to the time point closest to 8 weeks; if equidistant, we took the longer outcome)."

By no means does that say anything about the "long term" effects.