1 in 20: The life threatening bit
I am honestly thrilled and inspired by the response thus far to the 1 in 20 project. I am in the process of getting a web domain for it, and I will keep you posted.
One of the questions that I've frequently been asked about my little graphic is the "life-threatening" bit. Are all eating disorders life-threatening? And what does life-threatening even mean?
Here's my take. No, not everyone with an eating disorder is at immediate risk for dropping dead from physical complications or suicide. It also doesn't mean that most people with an eating disorder will die as a will die as a direct result of their disorder. Yet, when you look at the statistics, you see that someone with anorexia is 12 times more likely to die than the girl sitting at the desk next to her in class. This girl is also about 60 times more likely to attempt suicide.
IrishUp, one of my readers who I've been lucky enough to meet twice and on whom I have a massive brain crush, said this in a comment:
In medical terms "Life-threatening" is applied to any diagnosis, symptom, or condition where the risk mortality is considered high - particularly if untreated mortality is high.
This should not be confused with ALTEs - apparent life threatening events. These are situations where there is sudden respiratory arrest or obstruction, or acute severe cardiac or neurological abnormalities. One of the things that qualifies EDs as "life threatening" is that someone with active ED behaviors is at SUBSTANTIAL risk for ALTEs.
And let's remember that one of the real problems with EDs is that we (medical we as well as parental we) can't really tell when an ALTE is imminent. The way the body responds to the perterbations that ED behaviors subject it to, is to keep up homeostasis on vital functions for as long as possible, by pulling required resources from non-vital functions. And it keeps doing that until the biological reserve is used up. Which often doesn't show up on labs until the ALTE is already in process.
The acute elecrolyte imbalances from vomitting that can cause siezures and sudden cardiac death are another risk. You may not even be able to detect imminent events in a patient getting monitored daily.
LT is, of course, somewhat relative and subjective. When we are talking about somewhere between 0.6 - 1% mortality per year in a population whose expected mortality is 0.012%/yr, we are talking about GREATLY increased risk of death. When 10yr survival rates are less than that of (most) childhood leukemias, I think we can safely call EDs Life Threatening.
The fact that people don't react to ED the way they do to cancer is a huge part of the problem.
That's why I used the term "life-threatening." Too many people still view EDs as a choice or a phase, and they don't really understand that EDs are deadly serious.
13 comments:
Giddy-up girlfriend! I have been screaming at the top of my lungs about the seriousness of ED's. COUNTLESS numbers of people consider it a disease of immaturity or vanity. What it is IS a life-threatening disease. Not to mention QUALITY of life threatening ... e.g. loss of teeth, bones breaking, and of course all the psychological mess that comes from living in a world filled with food rituals.
I spoke to my tame Professor. She interestingly said that the new statistics just about to come out which include BED may make this number change considerably!
As soon as I can grab her attention long enough to find out more about her data, I will let you. In the meantime, Rock On, Carrie. xx
These numbers do include BED, actually. I'm guessing it's a bit of an underestimate, but it's the best data I could find.
How many studies are you basing this # on? From reading your last post, it sounded like one study, and I feel like that's worrisome in its lack of generalizability/reliability. I also worry about saying "every x minutes, a person is diagnosed" because it seemed like the study was looking at # carrying the diagnosis, not the rate of new diagnoses.
I think using faulty numbers to get people's attn about a disorder isn't right. It seems flashy and misleading.
I also disagree with the "life threatening bit" because it seems to suggest that another group of people may be diagnosed as having non-life threatening diagnoses. Further, I think we want to research/understand eating disorders not just because they're life threatening-- let's talk about how life-diminishing eating disorders are. Most people will not die of their eating disorders, but every person with an eating disorder will lose out on months/years of productive, healthy life.
Sorry this is really whiney, but as a fellow (soon to be physician-)scientist and recovered anorexic, I get fussy about this sort of thing. I worry that this is not going to help pediatricians, parents, or people with eating disorders unless it's very carefully researched and communicated more clearly.
JS, you said pretty much exactly what I was thinking. Thank you.
With all due respect, Carrie, I don't believe those numbers. The studies calculating SMRs (standard mortality ratios) vary a lot. Here are some things that can really skew it, that I feel your readers need to know:
- inpatient, outpatient, community sample, residential?
- BMI numbers, which is partially dependent on the sample too (inpatient/residential being worse than outpatient)
- sample size
- length of follow-up
- how they split up the groups: do they separate AN-BP and AN-R?
That's just SOME of the things. The mortality rates, that people have been quoting for a long time, come from studies of very sick, chronically sick patients. NOT THE MAJORITY OF THOSE WITH AN AND BN.
FOR EXAMPLE, here is data from an OUTPATIENT group: http://www.scienceofeds.org/2012/05/08/ednos-bulimia-nervosa-as-deadly-as-anorexia-in-outpatients/
THE SMRs are ~1.8, that's definitely not 12x more!
Taking a different approach and looking at a meta-analysis, which of course, has tons of caveats too (because it is hard to combine studies using different methodologies), suggests that mortality rates for BN and EDNOS are also around 1.8-1.9. And about 5.8 for AN. In this case, though, AN is chronic AN. Whereas most people do end up falling into EDNOS further on.
http://www.scienceofeds.org/2012/05/04/a-meta-analysis-of-mortality-in-eating-disorders/
Can you please provide the sources of the data that you are using? There are may studies on mortality rates, some suggesting SMR for AN is 30! But that's looking at girls with a BMI of <11.5!! NOT most people with AN.
I think the 1 in 20 number is misleading, and I think calling it life-threatening is more alarming that it needs to be.
Out of the deaths, 1 in 5 is suicide.
I'm sorry Carrie, I fully support the message you are trying to get across, and raising awareness about EDs. That's exactly what I want to do, too. But, as a science writer, don't you think it is important to explain the caveats and limitations of any particular study you pick (nevermind sourcing your work)? A lot of people read your blog, many, many more, than read mine. I think you owe it to them to be honest and transparent in your analysis and where the numbers come from.
Sincerely,
Tetyana
JS: I completely, whole-heartedly agree with everyone you wrote. Bang-on. Everyone needs to read your comment. It is misleading, and doesn't do anyone any favours.
My bias? I pick science education, science literacy and **critical thinking** skills, particularly when it comes to statistics and data (which people are awful at understanding, myself included, by the way) versus ED awareness. I do this everyday in the lab (completely non-ED field) and in my seminars. I get evaluated on critiquing peer-reviewed studies all the time, in class or during journal club. And so, I cringe when I see things like this. It is misleading to base a campaign of one study, with a relatively small sample size, in just one country.
Death due to suicide is an absolute *must include* in the mortality rate for ed's. Because it is a part of the brain shutting down after long periods of starvation or fluctuation in chemicals due to binging/starving cycles- it is part of the disease, just as strongly as death from heart attack or kidney failure.
Statistics are not an exact science. Never have been- no matter how carefully you measure a population, there will always be ouliers- and, of course, populations shift/grow/diminish over time. So to point a finger, wag it, and shout, "NOT EXACT!" is not very meaningful. Statistics never ARE exact.
They're just estimates, meant to provide some understanding or insight into a particular population. In this case, the stat "1 in 20" is meant to draw the reader into an open mind- that eds are more prevalent than they may realize- and the "life threatening" is meant to step into that open mind and show them that these eds are serious business, not societal fluff.
It's an AWARENESS CAMPAIGN, people. NOT a scientific dissertation.
yes, but I wouldnt want to be made aware of something that wasn't true. would you?
hm: SMR numbers always include suicides, they include all deaths. What I meant to say is that 1/5 of those dying due to their ED, are from suicides. And I agree, they absolutely must be included in the calculation, and they are. I agree that it is related to the disorder.
An awareness campaign should not be based on one study or false/elevated numbers.
You can say this is exactly what Naomi Wolf did, but to a lesser extent/less severe.
It doesn't benefit ANYONE to elevate the numbers. EDs are serious WITHOUT being life-threatening. My ED is not life-threatening, but it is has had a HUGE, HUGE, impact on my life and those around me. It doesn't have to be life-threatening to be important.
Believe me, I care about people understanding that it is serious. But why does that mean having to use dodgy numbers?
So, why not use 1 in 10? Or 1 in 5 if you want to grab the readers attention? Why stick to 1 in 20?
If you want to raise AWARENESS, you need to base it on credible science! Because if you don't, no one will take you seriously. At least no one that can go and look at the data, or anyone who can read more than one study and realize the numbers in the one used are likely outliers.
If you don't care to use good science for an awareness campaign, how can others trust you on the science about the genetics, or neurobiology, or psychology? How can others trust you as a science writer and science communicator? That's my problem with this.
As JS said, what's the point of being aware of something that isn't true?
Some good reads:
http://neuroskeptic.blogspot.ca/2009/05/questioning-one-in-four-part-1.html
http://neuroskeptic.blogspot.ca/2009/06/questioning-one-in-four-part-2.html
http://neuroskeptic.blogspot.ca/2009/04/help-theres-epidemic-of-anxiety.html
http://neuroskeptic.blogspot.ca/2011/11/one-in-four-revisited.html
I fully support an awareness campaign. 100%. I try to raise awareness about EDs on a daily basis. I am completely open about it to everyone who is interested in talking about it.
I think they are very serious problems. I think the quality of life and the health and psychological effects are very important to emphasize and, well, make people aware.
Don't get me wrong. This is exactly what I am trying to do with the blog, second only to general critical analysis of peer-reviewed literature and getting readers to understand that any one study has its strengths and limitations and biases, and it is important to know what they are.
So, I support Carrie's campaign idea. I'd love to put that on my blog, or share it on Facebook. But, for me to feel comfortable doing that, I need to feel like I can defend the numbers. And right now, I don't feel like I can.
That's all I'm trying to get across. Sigh. Sorry, I guess I just think it is important that if you go through with this campaign, you use good data to back up the very statistic that is the foundation of the campaign.
I hear you on wanting absolute truth in your facts. I get that.
But I don't think society can be pinned down to absolute specific numbers- because, as I stated before- society keeps changing- growing, dying, being born, being diagnosed, recovering. There IS no true absolute statistic, with regards to all of the population. Stats are always reductionistic. And awareness campaigns are even more so- because you have to come up with a simplistic stat that can grab the eye and the mind, and make a point. That simplistic stat and where it came from are pretty straightforward.
It's ok for it a campaign stat to be rounded and simplistic- that makes it understandable. And would it be so bad for it to catch the eye and make people WANT to research it to find out if it's *really* true? I think that's the point- to encourage the researching, the digging, the finding out of more.
I want to add one other thing, but I want to be careful/sensitive about it- saying one's ed isn't life-threatening, to me, sounds a bit like denial. ? Maybe I'm wrong, but it gives me pause- because eating is one of the core needs for life- and disordered eating of any kind throws sticks into the bicycle wheels of life. Maybe they're little twigs that just cause a bit of a bumpy ride. But it'll catch up, if you keep it up. Soon those wheels won't turn as smoothly as someone else's bike wheels. And then, you might fall off. Every ed is important. And every ed can be life-threatening- just because someone's life isn't in IMMEDIATE impending danger doesn't mean their overall lifespan isn't going to be impacted.
Thanks for the discussion! :)
Carrie, where are you??
Post a Comment