Science: Growing Too Fast?

There's a widespread perception among scientists that we're living in an era of relentless growth in terms of the number of scientific papers being published.

Many say that quantity has increased at the expense of quality: people are publishing "any old rubbish" or splitting their work into as many papers as possible, driven by the publish-or-perish culture of modern academia.

But is this true? To try and find out, I looked at the number of papers published each year, in English, on PubMed, for the past 30 years.

Here's the data: it shows an increase in the number of papers coming out each year, except for a small negative blip around the year 1997:

Now, when I first eyeballed this curve, I got the impression that growth has accelerated recently, consistent with the "recent pressure to publish" idea.

But here's the same data with each year's publications expressed as a ratio to the previous year's:

This reveals that the relative annual growth in the number of papers published has actually been pretty constant over the past 30 years. It's generally been around 4% (ratio of 1.04), and almost always within the range 2% to 6%. In other words, every year, scientists publish the same number of papers they did last year, plus about 4%.

The past few years have not seen especially strong growth, relatively speaking. At most we can say that year-on-year growth has been at the upper end of the historical range, 4 to 6%, but that's no faster than in the 1980s.

Still, this is a lot of growth. Assuming that it stays at 5% year on year, we'd expect a million new papers published in 2016, and two million in 2030.

But is that really feasible? Is there any good reason that science should grow exponentially in this way? Can that continue, or will we reach "peak science" or at least a plateau?

Obesity: Are We Food Obsessed?

According to a Professor Greg Whyte, writing in the Independent, when it comes to obesity, we've got an unhealthy obsession with diet. There is -

an incessant diatribe of diet propaganda purporting to possess the panacea for health... [but] the focus on diet linked to the volume and make-up of calories we consume has overshadowed the importance of the critical half of the energy balance equation: physical activity.

Clearly weight is, to a first approximation, a matter of calories in (diet) vs. calories out (physical activity). For any given diet, whether you lose or gain weight is determined by how much exercise you do, and vice versa. There's no such thing as "overeating" as such, there's just eating out of proportion to your level of exercise.

But have we forgotten that? Do we talk about the diet side of the equation more? I ran a few searches on PubMed and Google for "obesity" + various other terms to try and find out and it looks like Whyte is right.

See the graph above.

There does seem to be an imbalance, with "food" and "diet" being much more popular than "exercise" and "physical activity", both in terms of the scientific literature (PubMed), and more generally (Google). This is just a quick analysis of course, but it does suggest that when it comes to weight and obesity, we are more interested in calories in, than calories out.

I wonder why?

The Real Story On That "Antidepressant Surge"

Remember last week's story about how depression rates are soaring in Britain? It was all triggered by "new data" about an increase in antidepressant prescriptions.

At the time I was skeptical, not least because the data wasn't actually new, but I've done a bit more digging and it turns out the media coverage was even more misleading than I thought.

Here's some pretty graphs from the NHS Information Centre. I reiterate that all of these are freely available and have been for ages. Here's the one for antidepressants:

They've been rising strongly! In total prescription rates are about 60% higher now compared to in 2006. Oh dear.

What the papers didn't tell you is that pretty much every other class of drug has also increased over that period, by even more in some cases. Here's ADHD drugs, and dementia pills, which have increased by about 75% and 100% respectively:

There have also been steady increases in anticonvulsants and a 40% increase in meds for Parkinson's disease. All of the graphs are here.

So this suggests that there's been a general increase in prescriptions for brain drugs. But in fact it's even wider than that because if we look at the same data for cardiovascular system drugs, we find the same picture for most (although not all) kinds of these medications.

And for painkillers, we find over 50% increases in prescriptions of the stronger opioid drugs, a 20% increase in migraine drugs etc etc. I swear I'm not just copying and pasting the same graph.

Now clearly, all of these increased prescriptions don't mean that there are simultaneous explosions in rates of dementia, heart disease, pain, migraine, Parkinson's and ADHD, all in the past 5 years. We would have noticed if that were the case.

What's happened, clearly, is that doctors are just writing more prescriptions nowadays.

So it's misleading to say that there's been a spike in antidepressant prescriptions. Yes it's technically true but it ignores the context. The truth is that we seem to be experiencing a cultural shift in our relationship to medications - perhaps evidence of the creeping medicalization of life (although there are more prosaic explanations that need to be ruled out before we conclude that; this could be a bureaucratic change in the way prescriptions are counted.)

However, "Escalating Depression Crisis - Antidepressant Use Soars" is a better headline than "Possible Medicalization Gradually Continues For Sixth Year In Row".

The truth, sadly, has an inherent disadvantage in the battle for news coverage. If we find the truth boring, it's easy for someone to come along and make up something attention grabbing. But the only easy way to make the truth more interesting is to make it, well, less true.

Susan Greenfield Causes Autism

British neuroscientist Susan Greenfield has caused a storm with her suggestion that the recent rise in the use of the internet and social media may be related to the recent rise in autism.

I point to the increase in autism and I point to internet use. That's all. Establishing a causal relationship is very hard but there are trends out there that we must think about.

This has led to fellow Oxford neuroscientist Dorothy Bishop of BishopBlog writing an Open Letter asking her to "please, please, stop talking about autism". Twitter has been enlivened by #greenfieldism's such as "I point to the rise of Rebecca Black and the Greek sovereign debt crisis, that is all."

However, in a Neuroskeptic exclusive, I can reveal that the situation is far worse than anyone feared. Greenfield is not merely spreading unwarranted speculations about the recent rise in autism diagnoses.

She caused that rise.

The graph above shows the total number of scientific citations for Susan Greenfield's papers, over time. This is as good a measure as any of the influence Greenfield has had over our culture.

The trend is obvious, the growth is dramatic, and the correlation with the modern autism epidemic is undeniable.

Vaccines Cause Autism, Until You Look At The Data

According to a much-discussed new paper, vaccines may cause autism after all: A Positive Association found between Autism Prevalence and Childhood Vaccination uptake across the U.S. Population.

The author is Gayle DeLong, who "teaches international finance at Baruch College, City University of New York", according to her profile as a board member of anti-vaccine group SafeMinds. She correlated rates of coverage of the government recommended full set of vaccines in the 51 US states including Washington D.C., with registered rates of autism in those states six years later.

Uh-oh - there was a correlation between vaccination in two year kids, and the rate of autism in the state six years later, when those kids were eight. As the abstract says:

The higher the proportion of children receiving recommended vaccinations, the higher was the prevalence of AUT... The results suggest that although mercury has been removed from many vaccines, other culprits may link vaccines to autism. Further study into the relationship between vaccines and autism is warranted.

Sounds rather scary. Until you look at the data, helpfully provided in the paper. First up, here's the scatterplot of all of the vaccination rates and all of the autism-six-years-later rates:

There's more than 51 data points as you can see: there's actually 355 because each state had seven different datapoints (1995 vaccines vs 2001 autism though to 2001 vs 2007). This scatterplot shows no correlation. You can tell just from looking at it, but the correlation coefficient confirms this, as it's a tiny r 0.012 (from a possible range of 0 to 1).

To be fair, that's a very noisy measure, because each state has unique characteristics, so the effect of vaccines will be diluted. However, it's still a useful sanity check, and shows that there can't be a major effect, otherwise it would be too big to get diluted.

To get around this I next looked at the change in the rates of vaccination from one year to the next, and correlated that with the corresponding change in future rates of autism, within each state. A "change" of 1 means no change, 0.5 means it halved and 2 means it doubled, etc.

Zilch. Correlation coeffiencent r is 0.034.

Maybe the changes year-to-year were too small? So I checked the changes between the last year, and the first year.

This made the changes bigger, because more tends to change over six years than in just one. And, to be fair, this does produces a slightly stronger vaccine-autism effect... but it's still tiny. The correlation coefficient here, r, is 0.18 which means that vaccination changes accounts for 3% of the variability in autism changes (r^2 = 0.034.) The p value is 0.20, statistically insignificant.

My conclusion is that this dataset shows no evidence of any association. The author nonetheless found one. How? By doing some statistical wizardry.

The statistical model used took into consideration the unique characteristics of each state. For example, each state had a unique mixture of pollution, which may have affected the prevalence of autism, yet such an effect was not included in this study. A fixed-effects, within-group panel regression (Hall and Cummins 2005) controlled for these unique yet undefined characteristics by deriving a different starting point (intercept) for each state.

The 51 different intercepts - one for each state - reflected the base level of autism or speech disorders occurring in that state that were not explained by the other independent variables (vaccination rates, income, or ethnicity). The model then produced a single relationship between the independent variables and the prevalence of autism or speech disorders.

OK, that's all very fancy, but when the raw data shows zilch and you can only find a signal by "controlling for" stuff, alarm bells start ringing. Given sufficient statistical analysis you can make any data say anything you want.

If the author had given details of the methods, and explained why she chose to control for the variables she did, and not others, that might be different. But she didn't. Nor did she justify only looking at the effects six years later, when five or seven or ten would be just as sensible... and so on.

(Note: whenever I've said "autism", that's my shorthand for autism + SLI, which is what the paper looked at; autism alone data are not presented. Note also that by "vaccination %" I mean "% who got the full vaccine schedule"; the other kids may have got vaccines, just not all of them.)

Delong G (2011). A Positive Association found between Autism Prevalence and Childhood Vaccination uptake across the U.S. Population. Journal of toxicology and environmental health. Part A, 74 (14), 903-16 PMID: 21623535

Neurology vs Psychiatry

Neurology and psychiatry are related fields - if for no other reason, because neurological disorders can often manifest as, and get misdiagnosed as, psychiatric ones.

But what's the borderline between neurology and psychiatry? What makes one disease "neurological" and another "mental"? Are some psychiatric disorders more "neurological" than others?

It's a rather philosophical question and you could discuss it for as long as you wanted. Rather than doing that I thought I'd have a look to see which disorders are, at the moment, considered to fall into each category.

To do this I did a quick search the archives of two journals, Neurology which the world's leading journal of... well, guess, and the American Journal of Psychiatry. I looked to see how many papers from the past 20 years had either a Title or an Abstract which referred to various different diseases. You can see the results above. Note that the total number of papers varied, obviously, and I've only plotted the proportion.

Some interesting results. Schizophrenia, which is probably considered "the most neurological" psychiatric disorder, is in fact the least talked about in Neurology. Depression is top amongst the "core" psychiatric ones.

Autism occupies a middle ground, discussed by psychiatrists at 70% and neurologists at 30%. That didn't surprise me, but what did was that ADHD is almost as neurological as autism. Mental retardation is also intermediate, though it's 30:70 in favour of neurology. Whether autism is really less neurological than mental retardation, is a good question.

Then out of the disorders with a known neuropathology, Alzheimer's disease, Huntington's disease and "dementia" (which overlaps with Alzheimer's) are a bit psychiatric while stuff like headache and epilepsy is almost 100% neurological. Why this is, is not entirely clear, since both dementia and epilepsy are caused by neurological damage, and they can both cause "psychiatric" symptoms.

I suspect the difference is that it's just much harder to treat Alzheimer's, Huntington's and dementia. With epilepsy or meningitis, neurologists have a very good chance of controlling the symptoms and few patients will be left with ongoing psychiatric problems. But with the neurodegenerative disorders, neurologists can't really do much, leaving a large pool of people for psychiatrists to study.

Someone once said that neurologists take all of the curable diseases and leave psychiatrists with the ones they can't help. These figures suggest that there may be some truth in this.

British Government Fails Maths, Economics

The British government has decided to change the way English universities are funded. They say that this will improve teaching quality; I doubt it. Worse, however, is that the new scheme, the stated justification for which was to save money, is now seriously at risk of being too expensive. From next year universities will be allowed to charge up to £9,000 per year for undergraduate courses, up from the current limit of just over £3,000. However the government is also cutting their basic funding allowance by 80% to compensate. So government pays less and students pay more - eventually; the government will pay all the money up front in the form of a loan to the students.

The government's cost projections assumed that the mean fee at English universities, and hence the mean size of their loans, would be £7,500 per year. Why, no-one seems to know. Sources are unanimous that this was what they assumed, but no-one links to any kind of report explaining why. Maybe they gazed into a magic crystal ball. Parliament, performing a separate analysis, also worked under the assumption of £7,500, and their reason was that

we have assumed that... this fee covers the 80% reduction in [central government funding]. The average fee... is assumed to be £7,500 per annum for an undergraduate degree.

However, this is just silly. For averagefees to be £7,500, anyone charging some amount more than that, would have to be balanced out by someone charging the same amount less. That's what an average is.

However, no-one can afford to charge less, even if they wanted to, because they need to charge £7,500 to pay for their teaching and break even. £7,500 is the minimum not the average. But plenty will want to charge more. Oxford and Cambridge, for instance, were blatantly going to charge the top amount, because they're "top" universities. As a result, every other university which aspires to be elite will have to charge £9k, to keep up with Oxbridge.

Hence a domino effect goes down the line: every university will want to charge as much as the ones immediately ahead of them, so as not to look cheap. (The alternative, that they'd try to undercut them in price, makes no sense when you consider the amounts of money involved; the savings to the students would be minimal but the message - "we are cheap, therefore not very good" - would be loud and clear.)

I've whipped up a little plot showing all the universities which have currently announced their fees along with their position in the latest university rankings. A few small institutions are unranked and so don't appear.

The rankings go up to 115 so if the universities ranked over 58 charge over £7.5k, the others would have to charge less to cancel them out. I'll try to update this chart when fees are announced, but I think it's a forgone conclusion that this won't happen. Last updated 06/04/2011 10 am. See also here for a frequently-updated expert analysis.

The government is now seriously talking about having to cut what little direct university funding remains, in order to avoid losing money - from a policy which was supposed to save money. Yet this was always going to happen given what I said above. Indeed this policy, which was sold to the country as a cost-cutting measure, was always going to, at best, break even until the graduates repay their loans, and they won't even start doing that until the first batch graduate, in 2015 which is the next election year.

So there seem to be only two possible options. Maybe they knew it wouldn't save money, but in that case, why did they do it? It's not winning them any votes, so there must be a long-term plan, but what? The other possibility is that they genuinely thought it would save money. So it's a question of bungling incompetence vs. mysterious scheme. I'm not sure which is worse.

What Diseases Get Researched?

Dorothy Bishop (of BishopBlog) has a nice PLoS paper looking at: Which neurodevelopmental disorders get researched and why?.

She took 35 "neurodevelopmental" disorders, ranging from rare genetic syndromes like Rett's, up to autism, ADHD and specific language impairment (SLI), and compared their prevalence stated in a textbook, to the number of scientific papers published about them over the past 15 years.

Note that with something like Rett's, there's no question that they're problems with brain development. With autism, some people would contest that but not many nowadays. With ADHD and some others, however, it's pretty controversial. Bishop includes them on the grounds that they're generally treated as neurodevelopmental in the scientific literature.

The graph above - which I should stress is mine; Bishop's are much less messy - shows the basic results.

First up, there's a correlation between prevalence and the number of research publications, but as you can see, it's pretty weak. Within the rare genetic disorders (pretty much everything below 0.1% prevalence) there does seem to be a relationship. When you get to the more common disorders, which are also the ones which are more controversial, there's no correlation at all.

Some points stand out:

• Autism is very popular; it gets the same amount of research as intellectual disability aka mental retardation (ID/MR), even though ID/MR is 9 times more common (0.65% vs. 5.5%)
• Down's Syndrome gets a huge amount of research despite being rare. It gets much more than Cerebral Palsy and Fragile X despite them all being severe and roughly as common.
• Tourette's is much less studied than any other disorder with a similar prevalence.
  
• In the bottom left you'll see a bunch of apparantly very common disorders like dyslexia, dyscalculia, and specific language impairment, which are extremely under-studied... if you accept those prevalence figures.

Bishop also notes that both the number of publications, and the amount of US government funding, for ADHD and autism research have skyrocketed in recent years. Especially ADHD which had just 356 publications in 1985–1989 but this increased nearly twenty-fold to 6158 in 2005–2009!

As for why all these figures are they way they are, it's less clear. Bishop discusses various factors like severity and the availability of funding in the paper, but this can't explain everything. It seems likely that some things are just more scientifically fashionable than others, for whatever reason...

Link: See also Bishop's Guardian piece about the paper.

Bishop DV (2010). Which neurodevelopmental disorders get researched and why? PloS one, 5 (11) PMID: 21152085

The Rise of the Mouse

Everyone knows that scientists experiment on rats, and guinea-pigs. That's why we have "lab rats" and why, if you're trying out something new, you're a "human guinea-pig".

But this is all out of date. Nowadays, mice are the most popular lab animals. Here's a graph showing the number of scientific papers published each year, mentioning each kind of critter (data gathered with this script):

Rats were on top until about 10 years ago, when mice overtook them. Why? No-one wants to study mice if they can help it: they are horrible to work with compared to rats, and rats are more similar to humans physiologically. This is why rats were more popular for a long time. (Contrary to popular belief, guinea pigs were never used all that much, and they've become even less popular with the rise of mice.)

Non-scientists tend to think of rats as just big mice. They're not: mice are less intelligent, harder to handle (they bite... a lot), and they smell bad. The fact that they're smaller makes surgery, and even simple stuff like taking blood samples, much harder. On the plus side, you can fit more of them in any given space, making them cheaper, but that's about it.

So why did mice suddenly claim the crown? One word - knockout. Mice are the only mammal in which it's easy to perform genetic knockout, i.e. eliminating the function of a single gene. It's extremely difficult in rats, because, for reasons no-one really understands, it is harder to get rat stem cells to grow in vitro.

Knockout mice were "invented" in 1989, and the inexorable rise in the number of mouse papers began a few years later. Recently, there have been reports that knockout rats may now be easy; whether this will lead to a rat renaissance remains to be seen.

Knockouts have revolutionized biology, because they make it easy to investigate what each gene does. Just knock it out, and see what's wrong with your mouse. This is why there are mouse models of so many genetic diseases, while rat and monkey models are only available for a few disorders.

Are "Antipsychotics" Antipsychotics?

This is the question asked by Tilman Steinert & Martin Jandl in a letter to the journal Psychopharmacology.

They point out that in the past 20 years, the word "antipsychotic" has exploded in popularity. Less than 100 academic papers were published with that word in the title in 1990, but now it's over 600 per year.

The older term for the same drugs was "neuroleptics". This terminology, however, has slowly but surely fallen into disuse over the same time period.

To illustrate this they have a nice graph of PubMed hits. Neuroskeptic readers will be familiar with these as I have often posted my own and I recently wrote a bash script to harvest this data automatically. Now you too can be a historian of medicine from the comfort of your own home...

Why does it matter what we call them? A name is just a name, right? No, that's the problem. Actually, neuroleptic is just a name, because it doesn't mean anything. The term derives from the Greek "neuron", meaning... neuron, and "lambanō" meaning "to take hold of". However, no-one knows that unless they look it up on Wikipedia because it's just a name.

Antipsychotic, on the other hand, means something: it means they treat psychosis. But whether or not this is an accurate description of what "antipsychotics" actually do, is controversial. For one thing, these drugs are also used to treat many non-psychotic illnesses, like depression, and PTSD.

More fundamentally, it's not universally accepted that they have a direct anti-psychotic effect. All antipsychotics are powerful sedatives. There's a school of thought that says that this is in fact all they are, and rather than treating psychosis, they just sedate people until they stop being obviously psychotic.

Personally, I don't believe that, but that's not really the point: the point is that it's controversial, and calling them antipsychotics makes it hard to think about that controversy in a sensible way. To say that antipsychotics aren't actually antipsychotic is a contradiction in terms. To say they are antipsychotic is a tautology. Names shouldn't dictate the terms of a debate in that way. A name should just be a name.

The same point applies to more than just antipsychotics - I mean neuroleptics - of course. Perhaps the worst example is "antidepressants". Prozac, for example, is called an antidepressant. Implying that it treats depression.

But according to clinical trials, Prozac and other SSRIs are a lot more effective, relative to placebo, in obsessive-compulsive disorders (OCD) than they are in depression (though this is not necessarily true of all "antidepressants", yet more evidence that the word is unhelpful.)

So, as I asked in a previous post: "Are SSRIs actually antiobsessives that happen to be helpful in some cases of depression?" Personally, I think the only name for them which doesn't make any questionable assumptions, is simply 'SSRIs'.

Tilman Steinert and Martin Jandl (2010). Are antipsychotics antipsychotics? Psychopharmacology DOI: 10.1007/s00213-010-1927-3

The Fall of Freud

The works of Sigmund Freud were enormously influential in 20th century psychiatry, but they've now been reduced to little more than a fringe belief system. Armed with the latest version of my PubMed history script, and inspired by this classic gnxp post on the death of Marxism, postmodernism, and other stupid academic fads I decided to see how this happened.

As you can see, the number of published scientific papers related to Freud-y search terms like psychoanalytic has flat-lined for the past 50 years. That represents a serious collapse of influence, given the enormous expansion in the amount of research being published over this time.

Since 1960 the number of papers on schizophrenia has risen by a factor of 10 and anxiety by a factor of 80 (sic). The peak of Freud's fame was 1968, when almost as many papers referenced psychoanalytic (721) as did schizophrenia (989), and it was more than half as popular as antidepressants (1372). Today it's just 10% of either. Proportionally speaking, psychoanalysis has gone out with a whimper, though not a bang.

The rise of Cognitive Behavioral Therapy (CBT), however, is even more dramatic. From being almost unheard until the late 80's, it overtook psychoanalytic in 1993, and it's now more popular than antipsychotics and close on the heels of antidepressants.

What's going to happen in the future? If there is to be a struggle for influence it looks set to be fought between CBT and biological psychiatry, if only because they're pretty much the only games left in town. Yet one of the reasons behind CBT's widespread appeal is that it hasn't thus far overtly challenged biology, has adopted the methods of medicine (clinical trials etc.), and has presented itself as being useful as well as medication rather than instead of it.

One of the few exceptions was Richard Bentall's book Madness Explained (2003) in which he criticized psychiatry and presented a cognitive-behavioural alternative to orthodox biological theories of schizophrenia and bipolar disorder. Bentall remains on the radical wing of the CBT community but in the coming decades this kind of thing may become more common. Only time will tell...

How to Be A PubMed Historian

Quite a lot of people seem to like those graphs I sometimes make showing the number of papers published about a certain topic in any given year, based on the number of PubMed hits.

But how do I do it? Surely I don't sit there manually searching PubMed for each term, for each year, right? That would mean dozens, maybe hundreds, of manual searches. Well, unfortunately, that is exactly how I've done it in the past. I really am that cool, see.


Actually it doesn't take very long once you get into the swing of it, but I've now worked out a better way. See below for a bash script which repeatedly searches PubMed for a given sequence of years, downloads the first page of the results, picks out the bit where it tells you how many hits you got, and puts it all into a single output text file ready to be pasted into Excel or whatever. This comes with no guarantees whatsoever, but it seems to work. Enjoy...

Edit 29/06/2010: Vastly improved version that searches for multiple different terms sequentially, accepts terms that include spaces, and outputs the data into a sensible format. The search term text file should be a plain text file containing one search term per line. e.g:

serotonin depression
dopamine depression
GABA depression

Would search for each of those terms and output the data for each year into a single text file - with three data columns in this case - good for comparing the relative popularity of many different terms across time.

---
#! /bin/bash
# 29 . 06 . 2010
#PubMedHistory script by Neuroskeptic http://neuroskeptic.blogspot.com
# script to find out how many PubMed hits for a certain string in a given year range.

# usage: script (search term text file) (start year) (end year) (output file)
# e.g script list_of_terms.txt 2000 2005 dope.txt
#first, print the HEADER line of the output file.

printf "YEAR\t" > $4
cat $1 | while read subject
do
#pre-format the subject to remove spaces
ffa=${subject/' '/%20}
echo -n "$ffa" >> $4
printf "\t" >> $4
done
#and a newline
printf "\n" >> $4

#Now the real thing. The main loop is a YEAR loop:

for (( yearz=$2; yearz<=$3; yearz++ )) do #For each year, create a temporary file t.txt containing the output for this line.
#First, the year, then a tab.

printf "$yearz\t" > t.txt

#now, a second loop to go through the list of searches
cat $1 | while read subject
do
one=${subject/' '/%20}
wget -O $yearz.txt http://www.ncbi.nlm.nih.gov/sites/entrez?term="$one"+"$yearz"'[Publication Date]'

#find the line in the output with what we're interested in
output=`cat $yearz.txt | grep ncbi_resultcount | awk '{print}'`
#now, change it to get rid of the bit containing the search term
#as this will screw up the next step if it contains spaces!
output=${output/content*publication/LOL}
#print to a temp file
echo $output > temp$one$2$3$4.txt
#find the bit we want using awk
output=`awk '{ print $22 }' temp$one$2$3$4.txt`
rm temp$one$2$3$4.txt
rm $yearz.txt
#trim output
trimmedout=${output#content\=\"}
trimmedoutB=${trimmedout%\"}
#replace "false" with 0 because that's what "false" means
trimmedoutC=${trimmedoutB/'false'/0}
echo in year $yearz , I got $trimmedoutC. Saving to temp file t.txt
#write the result, and a tab, to the TEMPORARY output file
printf "$trimmedoutC\t" >> t.txt
done
#Now we've done all the search terms for this YEAR, so send the temporary data to the final file
cat t.txt >> $4
#and give it a newline
printf "\n" >> $4
done
rm t.txt

This Season's Hottest Brain Regions

Are you a budding neuroscientist who's not sure which part of the brain to specialize in? Or perhaps you're a purveyor of media neuro-nonsense who's wondering which area to namedrop as being the key to sex / intelligence / politics next?

Well, wonder no more, because Neuroskeptic can now exclusively reveal which parts of the brain are hot, and which are not, right now (thanks to the high-tech method of searching PubMed and counting the papers published referring to eight major brain regions, each year from 1985 to 2009.)

The hippocampus stands out as an extremely hot region with both a huge number of papers and rapid growth over 25 years. So it's probably a good place to build a career... but on the other hand, the market may be saturated already, and it shows some signs of flatlining in the past few years. The cerebellum has long been popular, but growth has been extremely slow lately.

To better highlight the growth curves here's the same data but normalized to the year 2000 (so "2" means twice as many papers as in 2000, etc.)

This shows major differences. The orbitofrontal cortex and cingulate cortex are both undergoing massive growth at the moment. The amygdala and parietal cortex are pretty hot too. By contrast, the cerebellum and the caudate are stuck in the scientific doldrums.

Why are the patterns so different for different parts of the brain? That's a big question which hopefully will get discussed in the Comments. I suspect that the recent rise of the cingulate cortex and the orbitofrontal cortex, however, has much to do with the rise of fMRI (i.e. within the last 10 years, mostly), which allows them to be easily studied in humans for the first time.

Both of these areas are quite difficult to study with older technologies like EEG, because of their location within the head. That said, the same problem applies to plenty of other regions, but the orbitofrontal and cingulate cortex are also difficult to study in lab rats and mice, because it's not clear which parts of the rodent brain map onto which parts of the human brain in these regions. By contrast, things like the cerebellum and caudate nucleus have exact rodent equivalents, perhaps making them more attractive to early researchers.

A Brief History of Bipolar Kids

Can children get bipolar disorder?

It depends who you ask. It's "controversial". Some say that, like schizophrenia, bipolar strikes in adolescence or after, and that pre-pubertal onset is extraordinarily rare. Others say that kids can be, and often are, bipolar, but their symptoms may differ from the ones seen in adults. You know a 20 year old's manic when they stay up for 3 days straight writing a book about how God's chosen them to save the world. A "bipolar" 10 year old, though, is more likely to show irritability and mood swings. Critics say that this isn't evidence of bipolar, it's evidence of... irritability and mood swings. Or, indeed, of being 10.

But what's not always appreciated is how new the concept of pediatric bipolar as a common disorder is, and how specific it is to American psychiatry. Here are a few graphs I put together to illustrate this, based on numbers of scientific publications.

First up, when did people start talking about it? Here's the number of PubMed hits for pediatric bipolar each year. As you can see, it was rarely talked about before the year 2000, after which its popularity shot up rapidly; it seems to have plateaued now, but it's hard to tell.

In fact, the true trend is even more dramatic, because many of the early hits were not about psychiatry at all. For example, in 1999, 5 of the 10 were nothing to do with manic-depression. One was about the growth pattern of a certain kind of bacteria (they're "bipolar", because they have two poles of growth.)

Is the post-2000 spike just a reflection of the fact that people are publishing more papers about bipolar in general? No. Here's a graph showing pediatric bipolar hits as a fraction of all "bipolar disorder" hits for that year. It's been rising for a while and it's now 5%.

Where are these publications coming from? America. Taking the first two pages of PubMed hits for pediatric bipolar, and excluding the non-psychiatric ones, 30 are from the USA, and just 4 are from elsewhere. For "bipolar disorder", it's 13 vs. 25. (This is in terms of the affiliation listed for the primary authors of the study.)

What about paediatric bipolar, the British spelling? It's almost unheard of. There are only 53 PubMed hits in total, as against 564 for pediatric bipolar. Of the first 20 hits, 9 are non-psychiatric, and 3 are from an Australian journal, criticizing the American concept of pediatric bipolar!

It's remarkable that the monthly British Journal of Psychiatry has never published a paper about "pediatric bipolar" or "paediatric bipolar": if you search their archives you get just 5 hits, and they are all in the references sections, not the papers themselves. The monthly American Journal of Psychiatry has published 37 papers mentioning "pediatric bipolar", of which 25 are not just in the references, and 10 are in the titles.

So, at least in terms of the literature, pediatric bipolar is overwhelmingly a 21st century American phenomenon. It barely existed before 2000, and it barely exists elsewhere. This corresponds to what some non-American psychiatrists have observed. In The Paediatric Bipolar Hypothesis: The View from Australia and New Zealand, Australian psychiatrists Peter Parry, Gareth Furber and Stephen Allison point out that

Traditionally, bipolar affective disorder has been considered rare in children and uncommon in adolescence ... However paediatric bipolar disorder (PBD) has become a topical issue in child and adolescent psychiatry over the last decade, driven by research in the USA. The proponents of PBD are concerned that the traditional approach to bipolar disorder in children and adolescents is missing a large number of distressed children, whose course of bipolar illness could be ameliorated or attenuated by early treatment.

Pediatric bipolar has certainly become more common as a diagnosis in the USA recently - a 40-fold increase in 12 years up to 2003:

The number of visits to primary care physicians in the under 20 age group where the diagnosis was bipolar disorder increased from 0.01% in 1994/5 to 0.44% in 2002/3

Whereas elsewhere, it's still regarded as incredibly uncommon...

Soutullo et al. reported that none of the 2,500 children 10 years or younger referred to the Royal Manchester Children's Hospital ... had a diagnosis of mania or bipolar disorder ... A more recent German survey revealed German child and adolescent psychiatrists were largely holding to a traditional stance as only 8% claimed to have diagnosed a pre-pubertal child with bipolar disorder.

Parry, Furber and Allison then present the results of a survey of 199 child and adolescent psychiatrists in Australia and New Zealand.

The majority of participants (53.4%) said they had never seen a case of pre-pubertal bipolar disorder, whilst a further 28.5% estimated they'd seen only 1 or 2 cases. Only 35 participants (18.2%) estimated having seen 3 or more cases of pre-pubertal bipolar disorder. ... Most participants (83.1%) were of the opinion that bipolar disorder in pre-pubertal children was either "very rare (less than 0.01%)", "rare (less than 0.1%)", or "cannot be diagnosed in this age group".

Of course this is just a survey, but the results are striking.

Peter Parry reports as a conflict-of-interest that he's a member of Healthy Skepticism, who are, in their own words, in the business of "Improving health by reducing harm from misleading drug promotion". I'm sure neither he nor I need to spell out why drug companies might conceivably have an interest in promoting the concept of pediatric bipolar disorder, given the wide range of drugs available for bipolar adults...

Parry, P., Furber, G., & Allison, S. (2009). The Paediatric Bipolar Hypothesis: The View from Australia and New Zealand Child and Adolescent Mental Health, 14 (3), 140-147 DOI: 10.1111/j.1475-3588.2008.00505.x

The Kids Are Alright

You may have heard about the amusing, er, debate between adult movie superstar Ron Jeremy and the video game industry:

Violent video games have "a much bigger negative influence on kids" than pornography, a leading porn star has claimed

Who's right? Neither. There are no big negative influences on today's kids, at least, none that have only recently started. Kids today are better behaved than they were 20 or 25 years ago, before any of the supposedly morally corrosive new technologies arrived to corrupt their minds: mobile phones, social networking, internet porn, violent video games...

Those are some strong claims I just made. The fear that something is very wrong in 21st century society, and that new technology has something to do with it, is widespread - whether the panic be about sexting, cyberbullying, the Facebook Generation, whatever - but the statistics tell a quite different and more positive story.

Crime rates fell, a lot, during the 1990s and have since declined a bit more, or stayed stable, in the USA (source):

In the UK, the dates are a little different but the recent drop is similar: crime rose up to the mid 1990s and then fell back down to where it started (source):

The picture is roughly the same in other industrialized countries. Bearing in mind that the vast majority of crime is committed by young people (specifically young men), this is evidence that something is not rotten in the state of today's yoof.

That's in terms of how they relate to others - what about how they feel about themselves? Have rates of mental illness increased? That's a difficult one because mental illness statistics are problematic, but in terms of the body count, suicide rates in young people have declined, albeit slightly, over the same period (source US, UK).

We don't know why crime rates fell. Everyone agrees that it happened, but everyone has their own ideas as to the cause, ranging from more abortions (the "Freakonomics theory"), to less lead pollution, to cellphones making it easier to report crimes, to... I'm sure you can make up your own. Ditto for suicide.

The point is, whatever reduced them, it's unlikely that something else was acting to increase them by any significant amount over the same period. It's possible - maybe something about 21st century life causes loads of crime and suicide, but luckily, some other mystery factor(s) reduced them even more at just the right time. But that's pretty implausible; if nothing else, Occam's razor tells us not to multiply explanatory factors unnecessarily. Which means it's implausible that the internet, video games, and the rest, are causing any significant degree of harm. Which is great news. Unless you're one of those pundits who loves bad news.

Mental Illness vs. Suicide

Do countries with more mental illness have more suicides?

At first glance, it seems as though the answer must be "yes". Although not all suicides are related to mental illness, unsurprisingly people with mental illness do have a much higher suicide rate than people without. So, all other things being equal, the rate of mental illness in a country should correlate with the suicide rate. Of course, all other things are not equal, and other factors might come into play such as the quality of mental health services. But it still seems as though there should be a correlation, albeit not a perfect one, between mental illness and suicide.

I decided to see whether or not there is such a correlation. The World Health Organization (WHO) provides the relevant data here. There have only ever been three studies attempting to measure rates of common mental illnesses internationally (1,2,3), and all three were run by the WHO. The WHO also collates national suicide rates (here) for most countries, although a few are missing. No-one seems to have published anything looking for a correlation between these two sets of numbers of before, or if they did, I've failed to find it.

So what's the story? Take a look -


In short, there's no correlation. The Pearson correlation (unweighted) r = 0.102, which is extremely low. As you can see, both mental illness and suicide rates vary greatly around the world, but there's no relationship. Japan has the second highest suicide rate, but one of the lowest rates of mental illnesses. The USA has the highest rate of mental illness, but a fairly low suicide rate. Brazil has the second highest level of mental illness but the second lowest occurrence of suicide.

*

Some technical notes: Two of the three surveys, the ICPE (2000) and the WMHS (2004), sampled the whole population of each country. The other one, which was also the earliest, the PPGHC (1993), surveyed people attending family doctors. Because this is a slightly different approach, I used the ICPE and the WMHS for the plot above, although the results from the PPGHC are very similar (see below).

The ICPE sampled 7 countries and the WMHS sampled 14, but 4 countries were included in both surveys, so there's a total of 17 countries. I've used the mean of the ICPE and the WMHS for those 4 countries where we have data from both, for the rest I've used whichever is available. For the suicide rates, the WHO gives data for various different years, so I've used 2002, or the nearest available year, since this is between 2000 and 2004. For two countries, Lebanon and Nigeria, the WHO do not report suicide rates. For China, rates of mental illness are given in both Beijing and Shanghai.

The studies used structured diagnostic interviews to try to measure the percentage of people suffering from mental illness in the 12 months before the interview. As I've said previously, this -

attempts to study a random sample of the population of a certain country. In order to establish whether each person is mentally ill or not, they use structured diagnostic interviews. These consists in asking the subject a fixed ("structured") series of questions, and declaring them to have a certain mental disorder if they answer "Yes" to a given number of them.

In this case the structured question interview was called the CIDI and it used DSM-IV criteria. You can check it out here. Example question:

You mentioned having periods that lasted several days or longer when you felt sad, empty, or depressed most of the day. During episodes of this sort, did you ever feel discouraged about how things were going in your life? (YES, NO, DON’T KNOW, REFUSED)

*

The rates from the population surveys (ICPE & WMHS) don't correlate with suicide but they do correlate with the rates from the PPGHC survey of people attending family doctors. The association here is very strong, with a correlation r = 0.693. The only outlier is the US. This is despite the fact that a decade elapsed between the first survey (1993) and the other two (2000, 2004).

This is important because it shows that the mental illness surveys are measuring something about these countries, something which is stable over time. They're not just producing random junk results. But whatever they're measuring, it's not related to suicide.

*

What does this mean? You leave a comment and tell me. But here's my take. I've often expressed skepticism of population surveys and their (very high) estimates of mental illness, and of the dubious political conclusions certain people have tried to draw from them, but even so, I was surprised to find no correlation at all with suicide. I'd say that any meaningful measure of mental illness should correlate with suicide. These surveys, using the CIDI, don't, so to me they're not meaningful.

One thing to bear in mind about these numbers is that they deal with "common" mental illnesses like depression, substance abuse and anxiety. They leave out the most severe disorders such as schizophrenia. Also, people in psychiatric hospitals, in prison, and the homeless, will not have been included in the studies because they sample "households". That could be why there's no association with suicide, but if so then these surveys are missing a very important aspect of mental health.

The surveys do seem to measure something, but I don't think it has much to do with mental illness. This is just a guess but I suspect they're measuring willingness to talk about your emotional life to strangers. At least stereotypically, the Chinese and the Japanese are known as more reserved in this regard than Brazilians and Americans. So it's no surprise that when you ask people a load of personal questions, the "rates of mental illness" seem to be lower in Japan than in America. This doesn't mean Americans are really more ill, just more open.

I've been talking about surveys looking at differences between countries, but if these are flawed, then so are surveys looking at just one country. For example, many studies have looked at mental illness in the USA using similar methods to these. But can we trust these methods bearing in mind that if you ask the same questions in, say, Belgium you get less than half the estimated rate despite it having double the number of suicides? Taken to its logical conclusion, maybe we know little about the prevalence of "common mental illness" anywhere.

Sartorius N, Ustün TB, Costa e Silva JA, Goldberg D, Lecrubier Y, Ormel J, Von Korff M, & Wittchen HU (1993). An international study of psychological problems in primary care. Preliminary report from the World Health Organization Collaborative Project on 'Psychological Problems in General Health Care'. Archives of general psychiatry, 50 (10), 819-24 PMID: 8215805

WHO (2000). Cross-national comparisons of the prevalences and correlates of mental disorders. WHO International Consortium in Psychiatric Epidemiology. Bulletin of the World Health Organization, 78 (4), 413-26 PMID: 10885160

Demyttenaere K, & et Al (2004). Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. JAMA, 291 (21), 2581-90 PMID: 15173149

Questioning One in Four: Part 1

Link - Part 2, Part 3

One in four people suffer mental illness at some point in their lives.

Everyone knows that. But where does that number come from? The answer may surprise. Join me, if you will, as I explore the biography of a statistic.

"1 in 4" is ubiquitous, at least in the English-speaking world. I can't think of another such number which is better known, except perhaps the fact that 1 in 3 people will suffer from cancer.

Anyone who's used the London Underground or watched British TV recently will be familiar with the Time to Change anti-stigma advertising drive. This £18 million campaign, run by the charities Mind and Rethink, is awash with "1 in 4"s, left right and center. Mind have it on their About Us page. The BBC have it on their main mental health page. There's even a One in Four magazine. And so on.

In the next post, I'll be examining the truth behind this statistic, but first, a little history. Google archive reveals that 1 in 4 is a child of the 1990s. English-language news media from the late 1980s contain the statement that in 1 in 4 (American) families will have a member who suffers from mental illness, but this is not the same thing.

As far as I can tell, "1 in 4 people" entered the popular mind in the early to mid 1990s. By 1995, it was common and being referred to as an accepted fact. See for example this snap-shot of the newspapers in 1995 under the search term ("one in four" + mental), showing that the idea had taken root by this point. Whereas the equivalent from 1992 is quite different.

Interestingly, the early 1990s also feature repeated references to 1 in 4 (Americans) suffering from mental illness in any given year; this statistic, however, gradually fades from view as the decade goes on. By 2000, 1 in 4 appears more often than ever, but now it refers almost mostly to lifetime prevalence.

These graphs show the number of Google archive hits from 1950 to 2008. I had hoped that this would illustrate my argument nicely, but sadly, the picture isn't all that clear. Here it is anyway - the top graph shows the increase in ("1 in 4" + mental) hits. The second shows, by way of comparison, the number of hits for just ("mental health"), which is much more level. That's nice. But the bottom graphs shows that ("1 in 8" + mental) also becomes more popular over about the same time-frame, which is a bit confusing, as 1 in 8 is not a number especially linked to mental health.

But - where did 1 in 4 come from? When I set out to write this post, I thought it would be fairly easy to find out, but having done a lot of digging, I genuinely don't know.

My first guess was that it must have been the National Comorbidity Survey (NCS). The NCS was an ambitious attempt to measure the prevalence of mental disorders in a representative sample of the U.S. population, masterminded by Harvard Prof. Ronald C. Kessler. Data collection took place between 1990 - 1992 and the results started to be published in 1993 - just about the time when 1 in 4 started to appear in the media.

But in fact the headline finding from the NCS, as published in 1994, was that the lifetime prevalence of mental disorders was nearly 50%! That's 1 in 2 (sic). The proportion estimated to suffer from a disorder in any given year was almost 1 in 3. But no sign of 1 in 4.

Meanwhile, in Britain, 1993 also saw the first Psychiatric Morbidity Survey, a similar enterprise. (Attentive Neuroskeptic fans will recall that this was the survey that the Mental Health Foundation recently distorted to make it look like rates of anxiety disorders are rising). Could this be the source? No, the headline number here was 1 in 6, which referred to mental illness in the past week, not over the lifetime.

Going further back, the Epidemiological Catchment Area (ECA) project, the first large-scale psychiatric epidemiology study, happened in the early 80's. The ECA famously concluded that 1 in 3 Americans suffer at least one mental illness over the lifetime, and 1 in 5 do in any given six month period! 3, 5 - but still not 4.


The World Health Organization quoted 1 in 4 lifetime in 2001, to much media fanfare, and I have seen the WHO given as a source for the figure. But where did they get it from? Well, good question.

Their report, New Understanding New Hope: The World Health Report 2001, notes that according to the WHO's own data, 450 million people worldwide currently suffer from a "neuropsychiatric conditions". With 6 billion people on Earth that's less than 1 in 12 (and that includes Alzheimer's, Parkinson's, epilepsy, etc.) And that's at any one time, not over the whole lifetime.

The report then quotes at least 1 in 4 as a lifetime prevalence (on page 23). Finally! But this is not based on WHO data. Instead, they cite three references: Regier et al. 1988; Wells et al. 1989; and Almeida-Filho et al. 1997. Let's check these references.

The first refers to an Epidemiological Catchment Area study of 12 month prevalence. Not lifetime. The ECA, as we've previously seen, gave a lifetime estimate of 1 in 3. The 12 month estimate is 15.4%, or 1 in 6. No 1 in 4 to be found here. The second refers to a 1989 paper from Christchurch, New Zealand. It reported a lifetime prevalence of 65.8% (sic) for any mental disorder. 2 in 3. For the "main" diagnoses, i.e. excluding most anxiety disorders, it was 36.6%. 1 in 3. The closest I could find to 1 in 4 in this study was 22.9% for main disorders, also excluding substance abuse disorders. 1 in 4, 1 in 3, or 2 in 3 - take your pick. The last reference is to a Brazilian study finding lifetime prevalence rates from 31.0% to 50.5% in three cities.

So, in 2001, the WHO quoted 1 in 4, but their only references, if taken seriously, put the lifetime prevalence is more like 1 in 2. So we still don't know where 1 in 4 comes from.

Recently, the National Comorbidity Survey Replication (NCS-R), another Kessler project, claimed a lifetime prevalence of any disorder in Americans of 50.8%. But the proportion suffering from a disorder in any one year was estimated at about one in four. So that's 1 in 4 at last, but that number appeared only appeared in 2005 - far too late to explain the origin of the meme. (And it was yearly, not lifetime, but you can see how people might have misinterpreted it.)

So, I give up. I don't believe there is a single source for 1 in 4. If anyone thinks they know where I've gone wrong, please let me know. But as far as I can see, 1 in 4 lifetime represents a kind of informal average of all of the studies I've discussed. It's a number that sticks in people's minds because it's high enough to capture the sense that "they're very common" while not being so high as to make people think "that's ridiculous" (as most of the actual estimates do). It's less a statistic, more a collective guess.

In the next post, I'll try to make sense of all these numbers.

[BPSDB]

Grant, B. (2006). About 26% of people in the US have an anxiety, mood, impulse control, or substance disorder Evidence-Based Mental Health, 9 (1), 27-27 DOI: 10.1136/ebmh.9.1.27