Tuesday, December 28, 2010

Back from break...Dr. Oz made me throw my remote at the tv! High heels CAUSE osteoarthritis...

Happy Holidays to all who are/have just celebrated. Mine was a whirlwind of activity with both my family and my husband's family joining us here in San Francisco.  After I said good-bye to my in-laws tonight, I settled onto my couch thinking, "I could use some mindless time." Lucky me.  There was Dr. Oz to wreck it for me.  The description on the listing was enough to make me want to skewer a lot of things. It began with artificial sweeteners, but I didn't make it far enough into the show without becoming enraged to watch it. Fear sells, kids. Fake causal relationships can stoke that fear...

The segment I am skewering has the title, "Do high heels cause arthritis?"  Naturally, for full effect, the headline was projected on a screen, on the back of his set with suitably ominous font.  The answer according to a researcher at Northwestern was  actually the answer to a different question. She did a study about the effects of heel height on knee joints and her findings are not altogether that surprising. Heels higher than 2" are correlated with damage to the knees. However, there is no way to say how often one would have to wear these to completely trash their knees. Fair enough. Next we had a podiatrist talk about a woman who was x-rayed while walking in heels of different heights. The 4" heels had her "out of alignment." Again not surprising. But then she said, keeping the body out of alignment CAUSES osteoarthritis. Is it a contributing factor? Perhaps.

CAUSES osteoarthritis? Give me a break. Apparently she seems to have forgotten her science training. And then they moved on to hocking shoe insets (one of which was made of flax seed - how many futures contracts on flax does OZ own?). I kid you not.

Buy this STUFF, if you can't give up your "ADDICTION" (yes, he used that very loaded word) to high heels. Otherwise, based on the CAUSAL relationship we FAILED to show, you WILL get osteoarthritis. I am off to have a glass of wine in 4" heels, without a flax-seed-padded insert. I'll be sure to report back when I have an osteoarthritis diagnosis.

Friday, December 17, 2010

As much as I wish it were true, Fox news makes you misinformed, not stupid.

"Study: Watching Fox News Makes you Stupid"

I love Business Insider, I must admit it. I have a big old Stat Girl middle-school-googlie-eyed-crush on Business Insider. That being said, since I'm a rational Stat Girl, I am afraid that I must be fair and skewer their reporting of studies too.
 The headline says, "stupid".   I don't know what study would ever say that. "Misinformed" does not mean stupid. It means being in possession and belief of information that is false or inaccurate. When you spew misinformation, you may SOUND stupid but your IQ hasn't necessarily changed.  If you didn't question your beliefs to begins with, well, that's another problem entirely.

What we find is that Mark Harwood's headline was even worse. It said, "Study Confirms That Fox News Makes You Stupid. He used "confirmed". GRRRRRRR. "Confirmed" and "causes" are as misused by writers as "I love you" is by that guy at the bar trying to get you to come home and sleep with him.

The study's flaw is not that it was run by a not-exactly apolitical organization, it is that it concludes that a causal relationship travels in one direction (let's not even get into whether there is a causal relationship at all). The HUGE problem with the study was that it tested the beliefs of FOX news viewers. It ASSUMES that their beliefs (as mistaken and idiotic as they may be) are CAUSED by watching Fox news. What if people with idiotic belifes are attracted to Fox news and watch it more and more? (Like depressed people taking mind-altering drugs). This is as likely as the reverse. Sorry kids, Fox news may have an audience more likely to be misinformed but we have NO right to say it's Fox's fault (AS MUCH AS I REALLY REALLY WANT TO).  Seriously, could a smart person really watch Fox news for more than a few seconds? Doubt it.
THERE IS A CORRELATION but that is not causality. And who knows if "stupid" or "watches Fox news" is the dependant variable in this case (then again, not sure it matters).

Thursday, December 16, 2010

Autism Research: Breakthrough Discovery on the Causes of Autism

I am not a fan of anecdotal evidence. To be clear, CAUSE is a very specific word and it has been used incorrectl­y and/or deceptivel­y in the title of this article.

Read the Article at HuffingtonPost

Autism Cause Identified! (no not really, just a bad headline). How to spot a nutbag...

Autism Research: Breakthrough Discovery on the Causes of Autism

This skewereing will explain how to spot a nutbag.

Step 1: Look for  for the word "CAUSE" where it does not belong.

Quackery, chicanery, or sell more booksism? Mark Hyman, a praciticing physician and therefore expert on autism (note not a neuroscientist, epidemiologist or peer-reviewed researcher) tells us that a recently published study in JAMA identifies the cause of autism as a mitochondrial disorder. And, what causes the mitochondrial disorder?...Mercury (queue the anti-vaccine wackadoos).   Help me please.

He used the word "CAUSES" in the headline. No one knows what casues autism. A potential mechanism for the disorder has been proposed. But this nutbag seems to KNOW the cause because he identified it in the anecdotal evidence of a little boy named Jackson.  Causality is notoriously difficult to prove. Is so difficult in fact that most people either prove a contradiction or disprove the opposite of something (Remember rejecting the null hypothesis?).
Step 2: Look for poignant anecdotes.
The story of Jackson is personal and interesting but it is an anecdote. This is a little boy saved by fish oil according to Dr.Hyman.  Maybe, who knows. Maybe the fish oil helped, maybe he was randomly "cured" at the same time he started taking fish oil. WHO KNOWS? The Huffington Post doesn't.

So, why is anecdotal evidence so powerful?  Actually, it goes to some psychology. There have been some tests done in behavioral economics that talk about why human beings are better equipped to help an individual (and are more moved to help an individual) than a group.  Dan Airley uses the example of the American Cancer Society using a network of cancer survivors to appeal to our sense of helping an individual. That is, one person who we know.  Look at Save the Children. They don't have you look at the staggering, heartbreaking numbers of starving children, they use one child.  Sadly, the quacks do this too. So do the anti-vaccine groups.

So what should we do? We MUST recognize that we are susceptible to the power and emotional pull of anecdotal evidence and return from "emotion mind" to "reason mind." If that fails you, look up pictures of individual children with polio or measles. That'll send you to the doctor for a vaccine, STAT.

Tuesday, December 14, 2010

Kids aren't "stuck" on sugary cereals ... or no kidding, this is a slow news week isn't it?

"Kids not so stuck on sugary breakfast cereals, study finds"

Oh Babycenter. Merry Christmas - turns out you don't have much to worry about with those diabetic-shock- inducing cereals after all.  Not so fast.  This study attempted to show that kids don't need sugary cereals to induce them to eat breakfast.  THIS IS NOT NEWS. If kids are hungry, they will probably eat what's in front of them. That is why what's in the house is probably more important than what's advertised on tv (opinion alert - that's my opinion).
This one is just plain old poor writing. George Orwell writes about the importance of clear, concise writing in his classic Why I Write.  He reminds us that ALL writing can be and is political to some extent. even if the writer is siply trying to express something, the politics of the writer will insert itself through word choice, sentence structure and the desire for the reader to identify with the author and his/her text.
End digression

The author of this headline seems to be saying that, children aren't addicted to sugary cereals. At least that's what the headline says to me. The study doesn't show that AT ALL. The study took 91 children (yes, 91 and they were mostly minority children in a summer camp setting - no amount of inductive reasoning can get you to any generalizations about children in the U.S. as a whole) and offered some of them a choice of "Fruit Loops, Frosted Flakes and Cocoa Pebbles." The other group were offered "Cheerios, Rice Krispies and Kellogg's Corn Flakes." Guess what? Both groups ate their breakfast. We didn't test whether or not, left to their own devices, the children chose the sugary cereals over the less sugary cereals. I think that test could have given us a headline that said the kids weren't stuck on sugary cereals. But, in this case, the design was bizarre. The children were given a set of choices where they could eat a sugary cereal or a sugary cereal. Likewise, they could chose a low sugar cereal or a low sugar cereal.

This is basically saying that kids will eat what's in front of them. If that's the conclusion, which was mind blowing for those " many parents [who] believe that if cereals aren't loaded with sweetness, kids won't eat them.," is the only advice, don't put sugar cereal in your child's bowl and your children will eat anyway? This is sad.  Not only did the methodology stink (sample was biased), also, the "test" and "control" groups didn't really answer a question with any useful information. The headline was deceitful because it tried entice someone to read an article about a terrible study with no REAL findings. Yay, Babycenter.com!!!!!! Everyday you add value. AS IF...

Thursday, December 9, 2010

I am "fit to be tied"... Cell phones harm the unborn? STOP THE BAD REPORTING!!!

The actually said that cellphone use could harm the unborn!!!

Babble.com, which purports to be for smart parents is full of it this time.  The article's title asks,
Is Your Cell Phone Hurting Your Unborn Child?. (As if the writer isn't trying to imply that it does). Anyone who remembers Alicia Silvestone in Clueless circa 1995 can imply my tone here.

Moms-to-be, all over the world just dropped their collective smartphones and recoiled in horror as this article told them that the conference call they just took is about to lead to behavioral problems for, WAIT FOR IT... THEIR UNBORN BABIES!!!!! This is disgusting. This is fear selling at its worst.  AAAAnd this my dears, is BS!

Here's where we find that the whole thing is a red herring, (below the fold, of course),

Researcher Leeka Kheifets says that the association between cell phone use and behavioral problems isn’t all that strong and, because the mothers were self-reporting, the data cannot be considered completely reliable.  However, she and her colleagues speculate that cell phone use might lead mothers to excessively secrete melatonin, which can impact her metabolism and potentially influence the brain development of the fetus.
At this point we must stop to consider that the RESEARCHER said that the association isn't all that strong. End scene. If it's not strong, why are we reporting on it?

Worse than that, they "speculate"s some causal power associated with melatonin.   Nice try. Not buying it. And neither should anyone. What is particularly damaging here is that the authors feed on a mother's rational desire to protect her child to create an irrational fear.

How to spot a study deigned to scare the heck out of you even though the conclusion is weak and/or non-existent:

as in "Is Your Cell Phone Hurting Your Unborn Baby?"
Adrenaline of the 100lb-mom-lifts-bus-off-child variet kicks in. "Yes. Mustn't it? If they are writing about it..."

Be assured the answer is more likely a relationship that is "not all that strong." Let me put it another way, "NO, there is not enough evidence to say anything about it..."
P.S. Don't get me started on the flawed nature of self-reported data...I'm just too tired to go there today.

Tuesday, December 7, 2010

Did You Have a Nice Day on the Right Wing, Dear? Fox Confirms the Married Men Are "Nicer"

 "Married Men Are Nicer and Here's Why"

Seriously.  This title actually confuses a correlation (which I think is not really that surprising anyway) with fewer occurrences of anti-social behavior and likelihood of a male twin to be married.  Ignoring the fact that this is not really a major aha moment (a##holes are less likely to marry) and therefore not really that interesting, the title assigns a CAUSE!! "Here's Why."  They think they know why!!! 289 pairs of twin men and Fox can tell you "why". As my eighteen-month-old is fond of saying, "No;no;no;no;no;no;no!" Envision a little blond head wagging back and forth. Also, is someone who engages in fewer anti-social behaviors "nicer"? What does "nice" even mean? The English language has many words in it. It certainly has enough words to be more specific and descriptive than "nice."

When anyone has a theory about why populations in a study behave differently, it is simply that, a theory. It is not the cause. It is not "the why." The writer in this case assigns an "explanation" to a set of data. In this case Fox decided that it was a data set in search of the family values crowd to explain what they believe they already know (marriage is good because it make people behave in a certain way). Give me a break!

It gets better, in a paragraph that begins with the incorrect usage of the word "however", the reporter goes even further. According to the study, for those pairs in which one of the twins married during the observation period, "anti-social" behavior decreased. They don't tell us by how much. They don't tell us how many were in the sample. There was a questionable control group (the ones who didn't marry). What if twins behave differently in marriage than non-twins and those men behave worse after marrying? We're reaching beyond the relevant range here.
Is this Fox's way of saying, "get those wild boys married, or anti-social behavior will destroy our country"? Give them a day to pass this on to some of their more insane commentators and it might be.

This headline is beyond ridicule. It is FALSE. We don't really know why. Incidences of anti-social behavior decreased after the men married. We also don't know if men with fewer anti-social tendencies to begin with are more likely to marry, or marriage made them less anti-social.  Honestly, I'm not sure that i care. One type of anti-social behavior was an episode of binge drinking. The difference between the two groups (1.3 occurrences and 0.7). So over the study period, the non-married men got drunk one extra time. Shocking. Well, not really.

One other thing, the men who married during the study were OLDER when they married thatn they were at the outset of the study. Anyone think that age and maturity might have something to do with a reduction in "anti-social" behavior? Just wondering.

We don't know why married men are "nicer" (or if they even "ARE" nicer). I repeat, WE DON'T KNOW WHY. And neither do you, Fox news.

Sunday, December 5, 2010

If you get divorced, your kids are 2x as likely to have a stroke...If you got divorced in the 1940s!!!!!

Struggling in your marriage? Here's HealthDay and Babycenter to warn you...
"Children of divorce face twice the risk of stroke as adults

Mon, Nov 22, 2010 (HealthDay News) — Children of divorce appear to have more than double the lifetime risk for experiencing a stroke compared with those whose parents' marriage stays intact during their childhood, new research suggests."
So, you're marriage is in trouble. You're having a rough time. Here comes Babycenter to make it that much easier to decide to stay in a marriage that you probably shouldn't or to feel so much guiltier and worse about what your decision to divorce is going to do to your children.  That's what you need right about now, isn't it?
When you read all the way down, you see that the people with double the occurrence were people of children who were divorced in...WAIT FOR IT..."1930s, 1940s and 1950s."  It's amazing to me that this is newsworthy. The study focused on Canadians in Manitoba and Saskatchewan. The results were presented at the annual meeting of the American GERENTOLOGICAL Association.  This is not for parents who might be divorcing today, this is for doctors of people over 60 who are trying to identify populations with stroke risk.  If we assume that the profile of a child of divorce in 1935 is the same as now, we are making an error. It even has a name! Extending the relevant range.
The best way to illustrate a relevant range is with water. Water, as it cools, becomes more dense.  So from 38 degrees to 4 degrees centigrade, water molecules come closer together. So let's just ASSUME that that relationship continues down to two degrees. We might, since we didn't measure the density of water at two degrees. But we would be wrong. Ever notice how ice floats on top of liquid water? That's because water, unlike other substances is the most dense at 3.98 degrees centigrade <http://www.helium.com/knowledge/6620-the-temperature-at-which-water-is-most-dense>.  
So if we assume that the children of divorces that occur in the 2010s will have the same increased likelihood for stroke as the children of divorce from the 1930s, 40s and 50s, we are looking outside of the relevant range. We can make NO assumptions about the likelihood of these children having a stroke over the course of their lifetime.

Long story short, Divorce is really difficult for everyone involved. It is a decision that I have anecdotal evidence telling me that most people don't make lightly.  If you are considering the effects of getting divorced on your children, PLEASE focus on things other than their risk of stroke in their sixties.  It's too far off and we don't know if it's correlated (don't get me started on causal). 

Friday, December 3, 2010

Why I Pick on Babycenter.com(not just because it's so easy)

A little about Stat Girl. I am the mother of a beautiful, wonderful 18-month-old daughter who has brought more joy, more challenge and more meaning to my life than I ever could have imagined. Also, I am a full-time outside of the home professional; a consultant who works in the financial field. I have an undergraduate degree in Philosophy (yes, you read that correctly, an employed Philosophy major - turns out there's a lot of us and we're pretty darn good too http://www.clemson.edu/caah/philosophy/information/famous.html). I have an MBA in Finance and Marketing. My brain did not fry from hormone overload when I gave birth (I was terrified that it might). More personal than all this is the fact I am in recovery from an Eating Disorder. As such, I take an SSRI every day. I took one every single day of my pregnancy and every single day that I breastfed. My daughter, according to a recent assessment, is ahead in almost all of her milestones except for "empathy" (I am smothering the snark about to rise from keyboard). When my daughter was 8 months old, I woke up from the "Do everything the experts say because you, as only a mere mother and not a baby professional, do not know what's best for your child". My BS meter started lighting up like a Christmas tree.

The catalyst was a post on Babycenter.com that was emailed to me in my weekly , "How to be a better mother" email from Babycenter. It talked about the "delayed onset of breastmilk in mothers who took SSRIs". That hit me in a more personal way than most headlines.  I read the article and it was flawed and DANGEROUS! Do you know what happens if you suggest to a fragile new mother, who is prone to depression (by the fact that she's on SSRIs to begin with) that she stop taking her pills? It's called Postpartum Depression and it's not pretty. So, I wrote to the kind folks at BabyCenter.com. Here's my note:

Dear Editors:
     I have to wonder whether or not your posting of content from the Health Day news service ismerely interesting and attention grabbing. It is perhaps ill advised. I have noticed that the health day reports rarely cite any concern over study methodologies. A particular example is the one pager suggesting that use of SSRIs is associated with delayed onset of breast milk.  Since so many mothers and expectant mothers are already weary of taking ANY medication for fear it will affect their child, the decision to continue on an anti-depressant medication is one that can be truly heart wrenching.  While the mothers and mothers to be know that they are already at increased risk for post partum depression and other complications, they also know that the effects of some of these medicines aren't known. In a population predisposed to depression and the concomitant symptoms of increased feelings of guilt and anxiety, they may decide to cease or reduce medication.  This study implies that a woman who takes SSRIs will have less success in breast feeding because the delayed onset of milk may require early formula feeding. Here is the rub, did this study control for maternal age at time of the birth, dosage of the SSRI, length of time on the SSRI, the version of the SSRI (some have much longer half lives than others? What about c section vs. vaginal birth? What about duration of labor and additional interventions? A HUGE influence on milk production is the mother's overall hydration level.  Were these mothers monitored for their hydration levels? Furthermore, what was the control group? How was the SSRI variable isolated?  Since none of these answers can be made in a short one pager, I must assume that, no, they probably weren't. Basically, this article will just make the decision to continue anti-depressant therapy (a decision that rightly belongs between a woman and the prescribing doctor when ALL factors of her situation are considered) even more difficult.  If ONE woman falls victim to post partum depression as a result of the reporting in this story, it is one too many.  Not ONCE in the article NOR in the "what can you do" section was the danger of discontinuing an anti-depressant without the supervision of a LICENSED medical professional mentioned. For a site that claims to be a place for mothers, I have to wonder how such a relevant issue went unmentioned.
Thank you,
Stat Girl (name edited, of course)
San Francisco, CA
Mother to Stat Baby (I'm actually going to start calling her that) - 8 months

Here's the response:
Hi Stat Girl,

Thank you for emailing BabyCenter with your thoughtful note.  You make some good points - I will forward your email to our news editors. Thank you for your interest in BabyCenter!

Community Administrator

NO FURTHER COMMUNICATION, EVER. Even after I wrote again:

         Thank you for forwarding my concerns. I am a bit trobled that there has been no additional response to my letter. Is it normally the policy of babycenter.com to simply reply "thanks for writing?". I have looked for any updates on PPD and have not seen much.  However, I have noticed more coming from the wire service Reuters (which is at least a reputable news source). I am encouraging the new mothers who I know to read the reporting of studies via babycenter.com with a high degree of skepticism.  Furthermore, in a survey that I was asked to take for babycenter.com, the survey seemed to be asking my degree of confidence in the information provided by the site.  If your goal and value proposition is to be "the authority on the web for all matters baby-related", I have to question whether or not your content team is aware of what such a goal entails.  If you are simply serving as a lead source for diapers.com, then I suppose catchy headlines are all that matters.  For example, today's email blast has a link to an article about VBAC. It says, "Vaginal birth after C section may be safe."  VBAC is VERY common. VBAC after multiple cesarians being safe, however is newsworthy. Making the lead link, "Vaginal birth after multiple C sections may be safe", seems like a more appropriate headline and does not imply that VBAC in general has heretofore been considered unsafe. I imagine that the click-through rate would be lower though.
Thank you again for reading my thoughts and passing them on.
Respectfully yours,
Stat Girl

MORAL OF THE STORY: This is my small way to save us from this kind of psychological warfare. That may seem hyperbolic, but that's how it feels to me.

Monday, November 29, 2010

More children have eating disorders than ever before, or, should we say more cases are being diagnosed and treated?

Happy Thanksgiving everyone! After all that food and fun, Healthday and Babycenter.com are back to remind us that when it comes to our children, now, more than ever there are more reasons to be terrified all day, everyday. Hyper-vigilance seems like the only answer. ENOUGH IS ENOUGH. This post will illustrate the difference in literal meaning of rate of incidence vs. rate of observation/diagnosis  (and it's a BIG one) and examine what I call the REALLY small denominator problem.

"Rate of eating disorders in kids keeps rising 

Eating disorders have risen steadily in children and teens over the last few decades, with some of the sharpest increases occurring in boys and minority youths, according to a new report."

My friends at Babycenter are back with a doosey. Eating disorders are among the most serious of mental illnesses and not to be trifled with. If you or a loved one are exhibiting any of the signs of an eating disorder as described here http://www.nimh.nih.gov/health/publications/eating-disorders/complete-index.shtml , seek help immediately. Go directly to a qualified professional; Do not pass go; Do not collect $200. 
Now that that's out of the way, back to the skewering.
According to this headline, more and more children have eating disorders every time we turn around. The evidence cited is a recent study that shows that , "hospitalizations for eating disorders jumped by 119 percent between 1999 and 2006 for kids younger than 12."

So the correct headline is "Frequency of hospitalizations for eating disorders in kids increased." Definitionally, more children may have eating disorders, but that's not what we see here. The other thing is that we don't really know if the rate of INCIDENCE of eating disorders is on the rise. All we DO know is that more cases are being diagnosed, documented and treated. It could even be that eating disorders are less prevalent than before, but because of better reporting and diagnostic tools, the rate of OBSERVATION is increasing. The rate of incidence implies that the absolute rate of eating disorders in the entire population is X. The rate of OBSERVATION refers the number that were counted and classified as an INCIDENCE of an eating disorders in the entire population. So if Susie had anorexia nervosa in 1998, but was untreated and undiagnosed, she still HAD anorexia nervosa. The INCIDENCE of anorexia nervosa was one. However, the number of observations for Susie would be zero. 

What we see here is that more children are being hospitalized for eating disorders. This may not even be a bad thing. What if, in the past, children with eating disorders were not hospitalized and instead were under-treated? Couldn't that mean that children were dying, carrying their disease longer into adolescence and adulthood, remaining in a sub-clinical state of misery for untold periods of time? Now that's awful. While hospitalizing anyone for anything is serious, not treating an illness that should be treated by hospitalization shows much more serious systemic problems (ignorance and neglect on the part of adults, mostly).  

On the other hand, maybe it's even worse. What if, in the past children were treated by mental health professionals and their pediatricians at an early phase and because of increasingly stringent and bizarre rules pertaining to health insurance plans limiting mental health benefits, children are having to reach death's door (literally) before they receive the treatment that they need? That's a different awful problem.

On the third hand (foot, I guess), what if these children were being hospitalized with a different primary diagnosis? Perhaps they have other illnesses such as depression, severe abdominal distress, or anything else for that matter and the eating disorder was not considered the primary cause of the hospitalization? 

And yet another foot, what if eating disorders weren't listed as the diagnosis code because some doctors knew that insurance would be more difficult to deal with?  According to someone I know, in the 1950's many doctors admitted Mrs. Smith with acute appendicitis and performed an appendectomy at the same time as an abortion. 

The truth is that we don't know if more children have eating disorders than in the past. We do know that more are being hospitalized at a young age with the primary reason for hospitalization being cited as an eating disorder.


The other disturbing part of this article is the idea that more and more boys and minorities are really seeing more eating disorders. Here we go again. Is it that these kids didn't have eating disorders, or that they went untreated and unreported? My guess is the latter, but I can only speculate because there is no base control to work form. But here's the more important lesson. the rate of anything occurring is a probability. That is a fraction reduced to 1 in X. A three fold increase sounds REALLY big but what if that is the increase from 1 in 1 Billion to 3 in 1 Billion? In the grand scheme of things, it's not a very significant change. It certainly wouldn't warrant a change in policy for the group that is 4x the population of the United States. (That's not to say those 3 cases aren't important, you aren't likely to come across the on or the three in your everyday life). 

In this article we hear that the sharpest increase is for boys and minorities. Let's do the math.

"...experts estimate that between 0.8 to 14 percent of Americans generally have at least some of the physical and psychological symptoms of an eating disorder, according to the report.
Boys now represent about 5 to 10 percent of those with eating disorders, although some research suggests that number may be even higher, said Lisa Lilenfeld, incoming president of the Eating Disorders Coalition for Research, Policy and Action in Washington, D.C."  
Ooh - catch that the first paragraph and the second are using 2 different data sets. Yikes (that's a problem for another post).
If math makes your eyes glaze over skip to the BIG bold part for a summary...
US population clock : 
For simplicity let's split the difference between 0.8 and 14 and say 7%...
310,814,706 * 7% = 21,757,029
21,757,029 * 10% (high-end estimate for boys) = 2,175,703
Let's just say that in the next year we saw 2,300,000 and the population stayed flat. the percentage increase would be (2,300,000-2,175,703)/2,175,703 = 5.7% increase.
Since non boys represent the rest of the eating disorders population is 21,757,029 * 90% = 19,581,326
If the same increase in CASES was seen there (about 124,297 CASES), we would see that there was an increase of (124,297/19,581,326) = 0.63%. 
5.7% vs. 0.63% HUGE difference but the same number of cases. So when you hear that the increase is sharper for boys this makes sense, they are starting off a much smaller base. 
If you sold one car today and two cars tomorrow, you had a 100% increase in the number of cars sold. The next day you sold three cars but you only have an increase of 50%, and so on... As the denominator increases the % change will continue to decrease.

Sunday, November 14, 2010

And we're back...Healthday tells us that social class affects your ability to get well if you have depression...

Please, someone hand me a frying pan to smack myself with. Maybe then I could lose enough brain cells to fail to notice the logic errors this wire service forces down our throats each day.

Here's the headline, 

Social Class May Affect Outcome of Depression Treatment

Really? How did Healthday get that from, 
Researchers reviewed the cases of 239 patients with major depression who took part in the U.S. National Institute of Mental Health Treatment of Depression Collaborative Research Program from 1982 to 1986.
The patients took antidepressants or received one of two different kinds of psychotherapy: interpersonal psychotherapy or cognitive-behavioral therapy. After treatment with drugs or psychotherapy, working-class and poor patients showed less improvement in their ability to function at work than did middle-class patients who had the same treatments, the University of Illinois researchers found.
1982-1986? Seriously. A meta analysis from OVER 20 years ago!!!  There were 3 groups. One group had antidepressants, one group had individual psychotherapy and 1 group had cbt.  Reminds me of "this little piggy, went to market...this little piggy stayed home."  Then, they tried to make conclusions about the efficacy based on the fact that "poor or working class" people having a different experience. Let's begin...
Sample size. How many of the patients were classified as "poor/working class" vs. "middle class"? Were they EQUALLY distributed in the three groups? What was the result for the "no treatment" control group ($20 says there wasn't one)? Each time you make the group smaller, it becomes harder to detect differences. Perhaps these things were addressed, Healthday doesn't tell us this, because Heathday seems not to respect our collective intelligence enough to do so.
Age of data.  Just so we're clear, antidepressant drugs have changed. Prozac and SSRIs came to market (there are those little piggies again) in 1987. Got that? A year AFTER the data were closed. The data may be irrelevant. Perhaps prozac would have worked wonders for both groups. Since those are the most prescribed class of antidepressants these days, it seems really silly to not have data about how patients do with that therapy. The title of the article would lead you to believe that drugs are less likely to help you when you're poor and depressed. Well, until someone adds the drugs we have today into the investigation, it's explanatory value remains limited at best.
Methods of treatment. No group had BOTH the drug and the psychotherapy. It is a well researched theory (NOTE I don't say FACT here because nothing in science can be "proven", even evolution remains a theory) that the most successful outcomes for depression patients come from a COMBINATION of psychotherapy and antidepressants.
Measurement of outcome. Success was considered "ability to function at work." Interesting. How the heck did they measure that? Supervisor interviews? Bonus payments? Performance appraisals? Attendance? Give me a break.
Oh and putting "may" in the title is a pathetic answer, so Healthday should give up the ghost and admit to writing for click through rates.
Please tell me what you think. Do you have a set of statistics driven journalism needing skewering? Let me know.

Saturday, November 13, 2010

Henry Blodgett Tells The Truth about the October employment numbers, or "This one goes to eleven..."

Today I must depart from my recent rants about medical journalism to give a shout out to Henry Blodgett and John Maudlin.  Maudlin writes a weekly investor email that parses economic data, reinterprets it, and offers far better analysis than most business journalists can muster in their entire career. Henry, well, he wrote about it and gave us a GREAT headline...

"Wow -- Check Out How Blatantly Our Government Misled Us With The October Jobs Numbers!"

Read more: http://www.businessinsider.com/wow-check-out-how-blatantly-our-government-misled-us-with-the-october-jobs-numbers-2010-11#comment-4cdf030ecadcbb55031d0000#ixzz15CUN8EMm

I encourage you all to read at least the Blodgett post, if not his source.

Basically, the government came out and said that we added 100,000 jobs in October. That's a nice story. The economy is on the mend, the bluebird of happiness is chirping away, babies are cooing and... wait, is that my BS alarm going off? Darn, I hate it when that happens.

Turns out, employment numbers are a result of a survey (mmmm sampling error smells good... but I won't even go there). After the survey folks at the BLS (and they ARE actually REALLY smart folks) apply a factor to the number to adjust for "seasonality".  What they are trying to accomplish is taking out the "noise" in the employment numbers. A perfect example is in June when all those young whipper-snappers graduate and can't find jobs and move home to mommy and daddy's basement.  No one wants to say, oh we now have 300K extra people without jobs (when actually, when you think about we totally do)... SO instead they "adjust out of that by multiplying the results by a factor of <insert result of scary looking algorithm designed to make your average mathphobe break into a cold sweat at the sight of it here>.

Well, they really did it this time.  They changed the factor. The following is a GROSS oversimplification for illustrative purposes. So 1,000,000 X .6 = 600,000. 600,000 employed people in September.  Now, the folks at the BLS thought this month seems different. All that Halloween candy made them giddy so they decided to change the factor to .7. In October, they say, 1,000,000 X .7 = 700,000. OK, but did we get 100K new jobs of course, not.

Here's the deal, anyone in government wanting to make a name for themselves would love to say 700,000-600,000 = 100,000 new jobs. IT'S A HALLOWEEN MIRACLE!

The reason this is so exciting is that a journalist has finally reported not on the number as released but on the manipulation of the data and how taking the number at face value is an insufficient form of inquiry.  

The other reason this is so funny is that the BLS basically just changed the scale of the ruler they use measure to measure employment. This one goes to 11... (If you don't recognize the reference, please add This Is Spinal Tap to your Netflix queue.).

Have a wonderful weekend. It's beautiful here in SF and I'm off to enjoy it with my husband and baby girl.


Friday, November 12, 2010

Don't blame accutane, the data sure as heck don't...

Before you make fun of me, "don't" in the title of this post is correct as the word "data" is the plural of "datum". Don't get me started on trying to figure out what, if anything, people mean when they use the phrase "data point". 

As hopefully now I have established, the purpose of this blog is to examine critically the ways in which statistics in the media are reported inaccurately (deliberately or out of ignorance) to the detriment of all of us. In today's Huffington Post, (I can feel a collective eye roll coming), via the Associated Press, we have a doozy of a headline.

"Acne Problems, Drug Treatments May Raise Suicide Risk

MARIA CHENG | 11/12/10 11:56 AM | AP"

A new study by Swedish researchers examined acne sufferers who took accutane (isotretinoin) to see if they had an increased incidence of suicide vs. the general population (remember increased incidence is NOT sufficient to imply causality).  They speculate (but DO NOT conclude) that it is not the drug, but rather the depression created form a painful and highly visible condition. The reason that they are required to speculate is that they could NOT reject the null hypothesis that there is "no relationship between suicide attempts and accutane." Rather than continuing to make a feeble conclusion that the data weren't (insert lame excuse here), they dealt honestly with the reality that until new evidence comes along, they cannot rely on the explanation that accutane wasn't more likely the CAUSE of any difference in the rate of suicide attempts between two populations than random chance.  The reason they couldn't was,
"Experts found the number of suicide attempts increased between about one and three years before the start of the treatment, though that increase was not statistically significant. The highest risk compared with the general population appeared to be within six months of the start of treatment. "

Since the headline is partially "Drug Treatments May Raise Suicide Risk", I have to wonder what gives? 
The author obtained a quote form another "expert." This doctor claims that all patients undergoing accutane therapy should be monitored for signs of depression. What we have here is one of the signs of propaganda I mentioned in an earlier post, http://statgirlskewer.blogspot.com/2010/11/and-on-to-breastfeeding-propaganda.html "2.Is the writer supporting the claim with anecdotal quotes from someone unrelated to the research? This is key because the author needs someone else to say what the ethical and scientific method trained researcher won't because it's dishonest."

Really, what we see is a statistically insignificant difference wherein, 1 more person in 2,300 (2 people in 2,300 - which is the difference of 0.04% and 0.09%) may attempt suicide. Naturally this is something that one may take into consideration when deciding if drug therapy for acne is right for them. Perhaps, candidates for treatment may need to be screened for depression prior to and during treatment. But to assume that an otherwise happy, well-adjusted person with acne will double their risk of attempting suicide because they took an acne drug would be illogical and irrational.

One thing that I liked about the researchers was that rather than trying to defend the idea that the drugs WERE somehow linked to suicidal behavior, even though the data said otherwise, was that they actually considered other reasons why this sample may have had even a slightly higher incidence of depression. Turns out, having a painful and visible illness might be enough after all, without blaming "big pharma."

Who knows? The point is the headline is MISLEADING and in order to get to the headline, the writer needed another doctor to cough up a quote.  Watch for my next post about necessary and sufficient conditions and why we need both together in order to draw rational conclusions.

Thursday, November 11, 2010

So now babies born via C Section are less aggressive? How to spot a theory looking for a study.

Here we go again. C-section. Does she? Doesn't she? It's dangerous. It's necessary. A c section is surgery and has risks associated with it.  But here's a new study that's noteworthy because it "suggests" that children born by c section are less aggressive. 
Without further ado, here is the headline,

"Childhood aggression linked to stressful birth

Wed, Nov 3, 2010 (Reuters) — Babies who undergo a difficult birth and are delivered using forceps are more likely to develop problems such as aggression during childhood compared with those born by Caesarean section, according to a study in China."

Later in the article Reuters tells us that researchers believe that elevated levels of cortisol are somehow related to aggression in children. To be clear, we are hearing a lot about cortisol these days, it's being blamed for people's inability to lose weight, inflammatory disorders and strange hives.  The basic understanding is that cortisol is a hormone released by stress.  

So in this article, the idea is that "cortisol, a hormone the body produces during a stressful and difficult birth." So, by "stressing the baby with forceps or vacuum extraction, you're dooming your child to aggression and "psychopathology".  


Cortisol comes from stress. Excessive stress, elevates cortisol levels. Kids who have social problems between 4 and 6 have elevated levels of cortisol.  More of these kids were delivered with interventions; ergo interventions at delivery cause social problems by elevating cortisol SOOO MUCH that 4 years later, the kids can't recover.  Feels like a stretch. No chance these kids have other sources of stress? Did anyone check for other things that they might have in common? IS it possible that they already have social problems and that is driving stress and heightened levels of cortisol? (the tail wagging the dog, so to speak). Of course it is. BUT THAT DOESN"T SELL ADVERTISING!

So kids who have high cortisol levels are more likely to be antisocial between 4 and 6. 

Well, maybe, but then again, maybe not.

According to, the Mayo Clinic staff,
"When you encounter perceived threats — a large dog barks at you during your morning walk, for instance — your hypothalamus, a tiny region at the base of your brain, sets off an alarm system in your body. Through a combination of nerve and hormonal signals, this system prompts your adrenal glands, located atop your kidneys, to release a surge of hormones, including adrenaline and cortisol."

Could these children live in noisier households? Could these children have stressful home or school situations? Do we really want mothers to believe that if their baby is delivered via forceps or vacuum extraction, she will cause mental problems later in life? As if she doesn't have enough to worry about getting ready to give birth.
Interventions at birth happen because of emergencies; not because they are planned. So here goes Reuters trying to scare the heck out of mothers-to-be and babycenter.com is ALL OVER IT! 

Tuesday, November 9, 2010

And on to the breastfeeding propaganda machine - How to spot propaganda by asking 3 questions

"Breastfeeding won't rob mom of sleep, study shows 
Mon, Nov 8, 2010 (HealthDay News) — It's commonly believed that one of the sacrifices new moms must make in order to breastfeed is their sleep. But new research suggests that's just not the case."

The study had 80 women in it - TOTAL! The purpose of the study was to squelch the idea that women use lack of sleep to justify formula feeding. Since formula feeding is tantamount to poisoning your child in today's La Leche League AAP run world of child-rearing, lots of research that supports the claim that breastfeeding is a) easy and b) desirable gets funding. What gets funded gets researched. The study showed that in a population of 80 women, the ones who breastfed got the same amount of sleep as the formula feeding mothers (in the first 12 weeks). It doesn't speak to the quality of the sleep, the help that women are or are not getting from relatives/husbands, etc. It does not address whether or not the mothers had reported sleep issues in the past. In short, there were not nearly enough controls to make any claim here. The null hypothesis that there was no difference in the sleep of the two groups (and there was a mixed group to screw things up further) was not able to be rejected. THIS PROVES NOTHING. It also doesn't "prove" that the women were getting the same amount of sleep, it was just saying that based on the variables measured and the size of the population (note n=80!!!) no difference could be detected that couldn't be explained by randomness. So there you have it ladies, even though the baby will probably sleep longer if formula fed, you may have less anxiety because when breastfeeding doesn't come easy for you, you stress out and feel like a depressed failure, you can't say that it was making you more tired. WHO ARE THESE PEOPLE? 

This, ladies, is mind control and propaganda pure and simple. Before you claim that I am anti-breastfeeding, I'm not. I did it. It was hard, painful and exhausting. I didn't feel like my life was made better by the experience, rather I felt trapped and scared. I continued for the good of my child. The day I stopped feeding my daughter "nature's perfect food" was when the pediatrician told me to go out and buy baby vitamins because breast milk didn't have enough iron (or DHA or whatever).  All that and still I had to buy manufactured nutrients for her. This went straight into my "You've got to be F'in kidding me" file.
At least the researcher stated that lack of rest was a factor in PPD. Did anyone check to see how many feedings the women were doing? If they were doing 90% of the feedings, guess what? It's because they're not getting any help which is not a factor that can be addressed by the decision to breastfeed or not. And that my dears is why you haven't had more than two consecutive hours of sleep in four weeks.
"Shelby Harris is director of the behavioral sleep medicine program at the Sleep-Wake Disorders Center at Montefiore Medical Center in New York City. She said: "Breastfed babies do wake up a little bit faster because they metabolize the milk faster, and it's always been assumed that moms wake up more, too. And, a lot of women give up on breastfeeding because they think formula will help them sleep more."
To be clear, Shelby Harris WAS NOT one of the researchers. What is "a lot" anyway? Let's start with the fact that she works in a sleep disorder center. Her frame of reference are women who are seeking the help of professionals for their sleep troubles. Perhaps they decide against breastfeeding at a higher rate than the general population. The fact is that the quote is unrelated, anecdotal evidence that doesn't belong in the article in the first place (paid by the word, weren't we?).


1. Is the issue "fraught with controversy" and/or supported by special interest groups on either side pouring money into the research?

2. Is the writer supporting the claim with anecdotal quotes from someone unrelated to the research? This is key because the author needs someone else to say what the ethical and scientific method trained researcher won't because it's dishonest.

3. Does it have a prescriptive quality? Ladies, we ALL already "know" that breastfeeding is best, so now let's fund research to take away your reasons for not wanting to rather than funding research to get you some darn help already...