SoylentNews is people
SoylentNews
A study analyzing data on over 1 million women has found an association between hormonal birth control and use/prescription of antidepressants:
"Today in vindication," wrote a woman on Twitter on Tuesday, summarizing the way many have received a striking new study [DOI: 10.1001/jamapsychiatry.2016.2387] [DX] that found those who use birth control — especially teenage girls — may be at a significantly higher risk of experiencing depression. The research, published in the Journal of the American Medical Association, has been heralded as groundbreaking for its depth and breadth, even if it seems to only confirm what some women say they've been feeling for years — that their hormonal contraceptives make them sad.
Women who used the combined birth control pill, a mix of estrogen and progestin, were 23 percent more likely to be prescribed anti-depressants than nonusers, and progestin-only pills raised the likelihood by 34 percent. With the patch, antidepressant use doubled; risk increased by 60 percent for vaginal rings and 40 percent for hormonal IUDs. And for teens age 15 to 19 taking combined oral contraceptives, the use of anti-depressants spiked 80 percent. Although those percentages may seem shocking, the absolute change is a small but significant spike. Among women who did not use hormonal birth control, an average of 1.7 out of 100 began taking anti-depressants in a given year. That rate increased to 2.2 out of 100 if the women took birth control.
It's the first study to conclude there might be a link between birth control and depression, author Øjvind Lidegaard told The Washington Post. Mood swings are often listed as a known side effect, but not clinical depression.
Also at The Guardian and CBC.
(Score: 5, Informative) by esperto123 on Thursday October 06 2016, @11:35AM
From what I can tell, this may only show that women that seek medical help from a gynecologist are more likely to seek help from a psychiatrist.
To actually show a link between hormonal contraceptives and depression it should be done a double blind and observe the ones on placebo having less depression or the one on contraceptives increasing the depression.
(Score: 1) by DECbot on Thursday October 06 2016, @02:57PM
I personally would be thrilled in the most sarcastic sense to discover that my wife's birth control was switched to a placebo.
cats~$ sudo chown -R us /home/base
(Score: 0, Disagree) by Anonymous Coward on Thursday October 06 2016, @03:40PM
In a double-blind study, neither your wife nor the experimenter would know it had been switched (nor would you).
(Score: 0) by Anonymous Coward on Thursday October 06 2016, @05:10PM
When birth control pills get switched to a placebo, you would know. A few weeks/months later, that is.
(Score: 2) by Arik on Friday October 07 2016, @12:00AM
YES!
Thank you for a not-stupid post. I love not-stupid posts.
To be fair it starts off at the beginning with appropriate reserve ("has found an association") but it certainly still comes off as endorsing the idea not just of an association but a 'link' with context clearly implying that the word is being used in a way that implies 'causal.'
Now your objection is one of the more obvious ones to make here, and imposing in it's undeniability, but it's obviously quite difficult, roughly impossible, to run properly controlled experiments on humans without colliding with generally accepted ethics policies. This is a problem for many fields, not just this one, and for as long as people have attempted to do science, not just recently. As a result, there is a commonly accepted work-around.
Didn't read this study to see if they did it or just guesstimated it and said it looks good (the latter is sadly common lately) but if they did it right (more on that in a minute) it would work something like this. They would not only plot the results according to the two main groupings, those on birth control and those not on birth control, but they would further break it down by all kinds of other variables, generally demographic - e.g. income range, education level, 'race' etc. They would then *adjust* the initial results based on that data. The idea being that at this stage you can detect and cancel signals caused by these other variables, leaving only the signal you are looking for.
Ok now the later part. Frankly I've always been skeptical of this. Yes you can control for certain well-understood biases this way, but it's dangerous to assume because a variable has been named and numbered it is actually understood. And of course the variables that are NOT named and numbered are clearly not being corrected for.
So there's a correlation. I'm pretty sure btw this was neither the first study to find the correlation nor the first to suggest there might be a causal relationship. I've read research on this before I feel certain.
Anyhow, there's correlation. But what does it mean? It could mean these methods of birth control cause depression, which is what the write-up is CLEARLY designed to imply.
It could also mean that depressed women are more likely to choose to take these methods of birth control.
Or, there might be no link between the two directly at all. There might be a third factor of some kind which causes women to be both more likely to take these method of birth control, and independently also more likely to be diagnosed with depression.
Oh, did you see what I did there? To be *diagnosed* with depression.
It's not like these people doing this 'study' actually went out and evaluated the girls using some objective gold standard and definitively determined each of the girls they have as a '1' in the depression field are actually clinically depressed. They didn't do that because that would be way too time consuming and expensive, perhaps, but even were that not a problem it's simply impossible to do. There is no gold standard test for clinical depression. One way to gain a little empathy, a little insight, into what clinical depression feels like is simply to contemplate how absurdly low the rate of agreement is when psychiatrists are double-blind tested on this diagnosis. But they're checking none of this, they're simply surveying who has the scrips and assuming they're all completely objective and dispositive.
And I suspect they know that, they probably even mentioned it, they probably think they gave themselve a big enough fudge factor to cover it. But the other side of the same assumption is less likely to have been apprehended. They are also, implicitly even if not consciously, assuming that the *lack of* a scrip for an anti-depressive drug indicates a lack of depression. To state such an assumption is to dispel it.
So I'll even go out on a limb here and offer a full alternative hypothesis. I suppose that women who take hormonal birth control are also women who are more likely than average to make a habit of going to the doctors - in many cases you have to go see a doctor to get them. And this would also neatly explain why they are more likely to have been diagnosed with clinical depression - because, similarly, you MUST go to a doctor to get this.