Believe or not, this blog post is about statistics, the human brain and common sense

UPDATE – The corresponding author of the paper, Dr. Suzana Herculano-Houzel kindly sent me a reprint of the paper. It was worth the wait; VERY interesting! After I read it more carefully, I will report…
(:-)… I love science!!!!!

Believe or not, this blog post is about statistics, the human brain and common sense.

I like statistics, I am just not very good at it (don’t you dare tell anyone!). Thank God for software packages! However, you may have the best statistical program and that is very good, but you must also have a reasonable working knowledge of what the statistics in an experiment mean, as well as a good measure of plain common sense to properly understand a given scientific work.

Not so long ago, a blog entry in The Guardian presented an excerpt of a 2009 paper*. The title of the blog post was: “How many neurons make a human brain? Billions fewer than we thought“, and in a subtitle it stated that we seem to be about 14 billion short.

The “top of the head” (pun absolutely intended) figure for the number of neurons in a normal human brain is 100 billion, reported in many textbooks, etc, but strangely enough the putative primary literature for this number is nowhere to be found. Fair enough. Now, I have not read the paper yet because it is behind a paywall. I sent an email to the senior author to try to get a reprint, but the information in the abstract (which is freely available) is enough to make my point for now, and I will of course update this post accordingly as soon as I read the paper. In the meantime, in Azevedo et al., they found that the average number of neurons in humans is close to 86 billion. From that perspective, yes indeed, we seem to be missing 14 billion neurons. Don’t panic!

The blog that reported these results said that the experiment was done by determining the number of neurons by an interesting cell counting method that I will not discuss here. Anyway, four brains were used, all from men, with an age range from 50 to 71 years. Here lies the first red flag; brains shrink with age; moreover, the rate of shrinking is surely affected by genetics and environmental factors. The second red flag: the sample number (4) is rather small.

Even more importantly, I am going ahead and predict what my wife will say about this post when she sees it tomorrow (she is peacefully sleeping by my side right now): “Of course they got a lower neuronal average, they used MALE brains!”

But I digress…

The third red flag: the abstract reports 86.1 +\- 8.1 billion neurons. This means that the actual number of neurons (the range) can be as little as 78 billion or as high as 94.2 billion (perilously close to the accepted figure, gasp!). Now, how significant are these results, statistically speaking? We do not know whether that value (the one after the +\-) represents the standard deviation, the standard error; the 95 % C.I. etc. We also do not know the p-value of the comparison against the 100 billion figure or the statistical method used (look these things up, they matter).

I want to clarify that I am not critisizing the paper. In fact, if find it quite interesting, that is why I want to read it!

Rather, I feel kind of annoyed when I read reports (like in the blog post) implied to be factual without the relevant information to reach a proper series of conclusions. It may well be that the 86b figure is significant and it tells us something about our brain, it may as well not, but we will not know which one it is with incomplete information.

The paper also presents a rather interesting observation about non-neuronal cells or glia. The “Common Wisdom” seems to be that glial cells are about 10 times as numerous as neurons. But in the paper and other related work, they seem to be found at roughly the same number as neurons. Now THAT’S interesting, and I hope to discuss it some other time.

In the meantime, I will keep using my neurons and try not to think about whether I have 80 or 100 billion of them.

The paper:
*Equal numbers of neuronal and nonneuronal cells make the human brain an isometrically scaled-up primate brain.
Azevedo FA, Carvalho LR, Grinberg LT, Farfel JM, Ferretti RE, Leite RE, Jacob Filho W, Lent R, Herculano-Houzel S.
J Comp Neurol. 2009 Apr 10;513(5):532-41.



TFB is available as an ebook (Kindle, Nook, as well as in iTunes). The hardcover is available at Amazon and at the Oxford University Press’ website.  There’s even a 20% discount code from OUP.  Shoot me a message if you have any questions!




You can subscribe to my blog! Just go to the “Home” page, right hand side.

Twitter: @Baldscientist



  1. The first thing I thought was standard error, then sample size, then how did they determine numbers. Why old men? Are they able to count neurons in alive brains or were they dead if so, how did they die? But the main issue is 86 different enough from 100 (billion) to be significant? Isn’t 100 billion rounded? So did 86 billion get rounded to 90 and if the method of counting neurons was such that it is all just a guesstimate, what the heck, just round it up to an even 100 billion.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s