transcript of episode 30: CLIMBING MOUNT THETA, 24th September 2022

[🎶 INTRO: "Spring Swing" by Dee Yan-Kee 🎶]

welcome to the error bar: putting the spod into brainsciencenewspodcast

in this episode: how to massively improve your memory in old age after just a few hours, why scientists won't change their mind, & how the brain is not made of light bulbs.

here is the brain news on the 24th September 2022:



the Independent reports that four days of brain stimulation can dramatically improve memory for words in older adults & the effects last for a month.

in a large & quite intensive study, a total of 150 adults aged 65 or over were asked to listen to lists of 20 words read aloud, then to repeat all those they could remember in the two minutes after the end of the list. they did this 5 times each day for 6 days. for all of the approximately 20 minutes that this took, they received electrical brain stimulation, of different kinds over one of two brain areas.

in the first experiment, a group of 20 adults had a high-frequency stimulation over the prefrontal cortex - an important part of the brain involved in thinking. & other things. another group had low-frequency stimulation over the parietal cortex - an important part of the brain involved in thinking. & other things. a third group had a fake kind of stimulation.

over the five days, as well as one month later, the high frequency frontal group were much better at remembering the words at the beginning of the list, while the low frequency parietal group were much better at remembering words towards the end of the list.

a second experiment reversed the stimulation types & found no improvements in memory. a third experiment repeated the first one & found similar results.


can electricity improve memory?

probably not.

oh, dear listener, the error bar laments.

i don't know how to deal with this kind of science. i can find nothing obviously or importantly wrong here, in the experimental design, analysis, or presentation of the results. indeed, it is very good. but i just don't believe them. they are too good to be true. i am in the 'uncanny valley' where data looks so good & realistic that there just must be something wrong.

there are some problems - there always are - but the problems i found are not serious. i list them here, because that's what i do:

first, the hypotheses, design & analysis seems to be based on a single paper published in 2013. nothing wrong with that, but the force with which the paper is written suggests there's a much-stronger evidence base than a single paper.

second, the style of the paper is SO 1990s - every piece of evidence is 'crucial' & 'decisive', there are 'gold standards', 'confirmations', 'rulings-out' & 'overwhelming majorities' in place of simple numerical facts, doubts or failures. everything seems to work exactly as it should have.

third, the results consist mostly of relatively-specific comparisons between individual groups, conditions, or time-points. that's fine, but there's no explanation of why some specific comparisons are given rather than others - they all seem a little arbitrary or selective.

fourth, it just seems unlikely that one brain area would do one task & be affected only by one frequency of electrical stimulation, while another brain area does another task & is affected only by another frequency of stimulation. it's just too neat.

but really i had nothing of substance here... until i remembered that the error bar covered a similar report from the same group way back in episode 2. let's have a listen...

"if true, this result would be an enormous and very specific effect of brain stimulation on behaviour. and that's the only reason i have for not really believing it - it's just too strong to be true..."

exactly! just eighteen months ago, this same group of researchers published some very clear & strong results using a similar method but on a very different topic - obsessive compulsive disorder. how can a group of researchers switch topics so quickly yet find - again - extremely-large effects of a brain stimulation method that many other researchers find extremely difficult to get working at all?


impressively dubious


the science was by Grover et al. 2022: Nature Neuroscience; reported in The Independent by @ninamasseypa on 22/Aug/22



every few months i am given the opportunity, via the academic channels that i paddle in, to scratch a favourite itch of mine, & ask: does theta-burst magnetic brain stimulation actually work?

this story follows-on nicely from the previous one about electrical brain stimulation, but with some important differences - while electrical brain stimulation doesn't generally work, some forms of magnetic brain stimulation do actually work. & that's not just my opinion, that's science.

but there are many different types of electrical & magnetic brain stimulation. one of the magnetic kinds is called theta-burst stimulation. it's a different but related kind of 'theta' to the 'theta' frequency used in other stimulation methods, including the one we just talked about in the previous story. in the magnetic case, theta means that a short burst of three electromagnetic pulses is repeated five times per second for forty seconds. this '5 times per second' is in the so-called 'theta' range of frequencies.

in 2005, a single study reported in the journal Neuron invented & then tested three different kinds of theta-burst stimulation on nine healthy human volunteers. they tested the effects of these three stimulations on how excitable the brain was afterwards. just like Goldilocks tasting porridge in the Three Bears' cabin in the woods, one kind of stimulation decreased excitability, one didn't do much at all & one increased excitability.

seventeen years later, this is exactly the story that most researchers still tell about this paper & this method of theta-burst stimulation. so, it was of course really very lucky indeed that the very first paper on the topic got it exactly right & collected enough data - from those nine participants - to show just exactly how this method must be interpreted for ever more. so it's great that we don't have to read any of the hundred or so follow-up papers.



do you detect some cynicism, reader?

i hope so.

the original Goldilocks paper has been cited by others about 3000 times. but a second paper, published by the same group of researchers eight years later & with six times as much data, has only been cited about 700 times. that's still pretty good! but lots of researchers are still ignoring it & repeating the Goldilocks story from 2005. why?

the reason, i imagine, is that the much better, more careful later paper did not find evidence to support the Goldilocks effects at all. instead, both of the forms of theta-burst stimulation tested did nothing, on average, across their 56 volunteers. awkward.

worse still, a meta-analysis of around 80 studies published a few more years later also found that the Goldilocks effects were: a) relatively small compared to the previous studies, b) affected by strong biases for researchers only to report data that supports the Goldilocks effects & c) very small or negative in the largest, most-precise studies.

that is not good news for the Goldilocks effect.

i've never used this method of theta-burst brain stimulation. one reason is that i don't have access easily to the equipment i'd need, but another is that it is just insane - 600 pulses of electromagnetism are sent into a person's head over a period of forty seconds. for that much electro-zapping, i want to know that it might, just might, actually work.

don't get me wrong: i am not saying that it doesn't work at all, or that there aren't good papers on it. no. i'm just saying we need to change the way we talk about theta-burst stimulation. it is not 'an established effect which is known to show blah blah blah'. rather, 'the effects of theta-burst stimulation are not currently well understood.'

i am telling you all of this because i recently saw yet another paper talking about 'individual variability' in the Goldilocks effect. i am amazed to watch, seemingly in slow motion, a whole research field continue to talk about how some people show the Goldilocks effect & some don't, & how that's really very interesting indeed & we need to know more about it. this is 10 long years after the discoverers of the effect showed so beautifully how it categorically does not work the way they first said it did.

theta-burst researchers: it's time to admit you have a problem.


give it up while you can


the science was by Park 2022: Journal of Neurophysiology, & Chung et al. 2016: Neuroscience and Biobehavioral Reviews;

and the brain in brief...



the title editors at the New Scientist write that a new study finds which brain areas "light up" when we look at food. the areas are of course: none of them, because the brain is not made of light bulbs. it's really easy to tell this because when you close your eyes, it is dark inside your head.


don't say 'lighting up'


reported in The New Scientist by Jason Arunn Murugesu on 19/Sep/22



this year the university of Oxford lost two long-serving professors of vision science, both born in 1944.

Sir Colin Blakemore, vision scientist, neuroscientist & science advocate died in June 2022. his influence is too great to cover here, so here are two anecdotes:

Molnár & Parker in Current Biology note that, as Oxford's youngest ever Waynflete Professor, "When he showed up for his first faculty meeting at Oxford, the chair told him that “student representatives are not required for these items”".

during my PhD, i attended a seminar by a visiting academic who promised to use brain imaging to distinguish between the 'analytic' & 'synthetic' statements described by the eighteenth century German philosopher Immanuel Kant. the small seminar room was packed. the academic presented his brain imaging experiment, which involved people reading long sentences in the synthetic condition & short statements in the analytic condition. comparing synthetic with analytic resulted in greater responses in the middle of the frontal lobe, just in front of the main brain area for movement. from the back of the room, Professor Colin Blakemore said only "Now that's frontal eye fields." & it was left for others to explain later that people need to make more eye movements to read long sentences than short statements. to this day, the neural correlates of Kant's philosophical system remain undiscovered.

a second loss for the University of Oxford is Professor Oliver Braddick, who died in January & whose obituary appeared recently in the journal Perception. a life's work in the development of the visual system, he was head of Experimental Psychology in Oxford. apparently great at pub quizzes, according to his obituary "his son describes him as ‘Google before Google was invented’.


rest in peace


the science was by Philipp-Muller et al. 2022: Proceedings of the National Academy of Sciences of the United States of America, Greer & Firestein 2022: Nature Neuroscience, Molnár & Parker 2022: Current Biology, Molnár & Hannan 2022: Nature Neuroscience, Charman 2022: Ophthalmic and Physiological Optics, & Perception 2022: Perception;

[🎶 OUTRO: "Cosmopolitan - Margarita - Bellini"by Dee Yan-Kee 🎶]

it's closing time at the error bar, but do drop in next time for more brain news, fact-checking & neuro-opinions. take care.

the error bar was devised & produced by Dr Nick Holmes from University of Birmingham's School of Sport, Exercise and Rehabilitation Sciences. the music by Dee Yan-Kee is available from the free music archive. find us at the error bar dot com, on twitter at bar error, or email talk at the error bar dot com