transcript of episode 1: GREEN PATHS TO BRAIN HEALTH, 15th January 2021
welcome to the error bar: where neuroscience meets uncertainty
in this episode: the vegetables & vitamins that stop cognitive decline & parkinson's, how city life slows you down, & the continuing failure of computers to be like the brain
here is the brain news on the 15th January 2021:
[🎶 BIG BEN BONGS 🎶]
VEGETABLES FOR BRAIN HEALTH
a pair of papers claim that brain degeneration can be prevented by the foods we eat.
in the first paper - which was not freely available online - the claim is that vitamins c and e found in peppers and broccoli help prevent the degenerative disease parkinsons's disease.
the second paper looks at the historical food reports of 10000 people in the USA, along with how they performed on follow-up cognitive tests every three years.
only half of the 10000 volunteers' data were used - some had died, and others hadn't filled in the food questionnaires correctly. somewhat suspiciously, people who reported eating too little or too much food, or had poor mental state scores, were removed from the analysis.
the headline-grabbing claim is that a mediterranean-style diet with lots of fruit and vegetables is only good for your brain if you also don't eat lots of western food on the side
[🎶 INTERLUDE 🎶]
when exactly are vegetables good for your brain?
the claim is that mediterranean food is only good for you if you don't also eat lots of western food. but that's not actually what the study found. it's more subtle and contextual than that.
as is common in studies like this, there are lots of possible variables to take into account - the age, sex, ethnic group, and pre-existing health conditions. for example, people with a high western style diet were more male, and those with a high mediterranean diet also took more exercise.
but let's just examine the central claim. if you eat lots of mediterranean food, that is always a good thing - cognitive test scores were higher and declined less slowly. and if you eat lots of western food, then you were a little worse off overall. but these two trends were totally unrelated to each other - not what the headline claims.
the researchers then divided the western diets into a 'high' and a 'low' diet and did analyses on each part separately. this 'median split' is a common tactic - divide the data into slices and look at each slice separately.
in the low-western slice they found slightly stronger relationships than in the high western slice. but the direction and size of these relationships was quite comparable. so, while the conclusions were that a mediterranean diet is only good for you if you don't also eat lots of western food, actually, the benefit of eating lots of mediterranean foods was always about 6 years' additional brain health, for both high and low western diets
eat a balanced diet, lots of fresh fruit and vegetables, exercise lots, stay mentally active; don't believe everything you read in retrospective food questionnaire studies
the science was by Agarwal et al. in Alzheimer's and Dementia, & Hantikainen et al. in Neurology; reported in The Daily Mail by @RyanMorrisonJer on 7/Jan/21, & The Daily Mail by Stacy Liberatore on 11/Jan/21
[🎶 BIG BEN BONGS 🎶]
URBAN LANDSCAPES SLOW YOU DOWN
in a long room with a large video screen at one end, 20 adults were asked to walk over a 10 metre course while looking at a picture. they did this 105 times - once each for 100 scenes - 50 urban and 50 rural - as well as 5 times with a blank grey scene.
the researchers recorded the positions of various parts of the volunteers' bodies - ankles, knees, and hips, and measured walking speed, step length, step time and leg swing time.
they found that when watching the urban scenes, people walked slowest - at 2.88 miles per hour while rural scenes were walked towards at 2.91 mph. the results are interpreted in terms of the visual 'stress' or 'distraction' of looking at urban landscapes.
[🎶 INTERLUDE 🎶]
do urban landscapes slow you down?
while the slowest walking was indeed produced by the urban scenes, this was only very slightly slower than the rural scenes - by 0.03 miles per hour, or a difference of only 60 metres after walking for a whole hour.
is this a big difference? not really. is it a statistically reliable difference? yes - the data and the analysis were powerful enough to find statistical differences, but these differences were not as strong as the control study reported in the supplementary materials, in which the volunteers walked along while doing some cognitively demanding stuff.
most importantly, while the differences between urban and rural scenes were rather small, the differences between either of those conditions and a blank grey scene were much larger. walking towards neutral grey images was 0.1 mph faster than towards a rural scene - that's 150 meters extra per hour.
watching cityscapes may slow you down a bit, but watching a blank screen really puts a spring in your step
and the brain in brief...
[🎶 BIG BEN BONGS 🎶]
AT LAST! COMPUTERS STILL CAN'T SEE
some new processors in some new televisions are described as intelligent because they adapt what is being shown and played to the viewer, depending on what else is happening on other parts of the screen or on the audio.
[🎶 INTERLUDE 🎶]
are they really 'cognitively intelligent'?
it may very well be clever, but it isn't human-like artificial intelligence. the scientific problems of computer vision and hearing run a little bit deeper than the latest televisual gadget
An introduction to the error bar
Hello! welcome to the error bar!
my name's Nick Holmes, I'm a scientist and lecturer at the University of Nottingham in the UK, working in the psychology department, studying touch and movement in adults and children.
I hope you enjoyed the brain news
in this first episode I want to tell you more about why the error bar exists, and what you can expect to hear over the coming weeks and months.
if you don't want to listen to the rest of this introduction, then the error bar is a podcast for scientists, journalists, and the general public who want a quick, up-to-date opinion from a card-carrying neuroscientist to help them de-cipher news reports about the brain.
why does the error bar exist?
in the first few weeks of the Great Coronavirus Pandemic of 2020, I was reading a coffee table magazine and it said: 'Whatever you do during lockdown, DO NOT START A PODCAST. You'll regret it, you won't have time, the novelty will wear off. This is not you, Nick ''I'm pretty sure it used my first name - We don't need another middle aged man blah blah blah...'' I sort of tuned out at that point, but what I do remember is the strong advice about definitely starting a podcast.
i had already been thinking about doing a podcast, I bought the error bar domain name early in 2020, and I thought that fact-checking brain science in a Post-Truth age of Fake News and Alternative Facts was a thing worth doing. so if all I do is prove the Coffee Table Magazine wrong, it will have been worth it.
i've been listening to podcasts a lot, when driving, taking a shower, or digging holes in the garden. Podcast reviews are now appearing in The Week magazine, everyone on Twitter is a host or a guest of something. Although a relatively late adopter, I think I can do a fair job of being as fabulous as Hannah Fry, as exciting as Adam Rutherford, as learned as Jim Al-Khalili and as numerate as Tim Harford. And if I can't, then it's likely going to be down to laziness. And talent.
what is the error bar?
the error bar aims in part to be an electronic version of some of interesting scientific conversations I've had. these are almost exclusively in a pub or bar, talking with friendly scientists over a pint. we'd chat about scientific results that aren't what they seem, or methods that just don't work, or common statistical problems, or poor interpretations of science that make it into the public domain and need to be fact-checked. the error bar is here for that.
a few weeks ago on twitter I saw a post saying: ''It's the media's responsibility to report science accurately.'' yeah, it is. but journalists aren't all scientists or statisticians; they might work only from a press release, or a commentary or review in another magazine that they have access to; they might be under a deadline that doesn't suit reading a whole scientific paper. then the journalists or editors or headline writers may have mis-understood or mis-represented the science.
the British press, and the tabloids in particular, have a lot to answer for - politically and socially at least - and part of me worries that if there is no alternative scientific narrative or direct challenge to the tabloid versions of brain science stories in the press, then are we going to find ourselves, 40 years from now, leaving science because the public voted it out'
so the error bar is here to read the neuroscience papers behind the news, and to give a quick, semi-expert opinion, with no gloss.
how it's going to work
every two weeks at first, I'm going to record a 9-minute summary of my interpretation of the research papers behind the news stories in the previous 2 weeks. the first news bulletin came today ' and there's a pilot episode available from April 2020. I've linked the podcast up to a website theerrorbar.com, and to a twitter account ' @BarError, so you can follow us there, with all the episodes, stories, scripts and links to further reading. the website and twitter will repeat all the stories on the podcast, plus a few more, they'll give a full transcript of the spoken words, and a bot that I've programmed will send a regular series of tweets about the show the stories and papers.
once I've got the news bulletins, website, and twitter working smoothly, then, about once a month, there will be a more in-depth discussion or interview aimed at delving a little deeper into one of the stories that we covered, or into something that's relevant or timely, or just interesting. we'll focus as much as possible on news, mistakes and controversies.
other podcasts are available
the market for neuroscience podcasts may already be saturated, so we'll see how this goes. The main aim of the error bar is to provide rapid-fire fact-checking of news stories, plus quick discussions of timely controversies. There are some great neuroscience podcasts out there for those wanting longer more in-depth discussions about the wider world of neuroscience, and I'd recommend you look at Everything Hertz; Two Psychologists, Four Beers; WaterCooler Neuroscience. the highly successful EverythingHertz once said that their show was ''unprofessional and uninformed'', which I think means that there is market space for at least three more podcasts, professional, informed or otherwise.
in my mind, the error bar will be a cross between More or Less, Talking Politics, and The Day Today. we'll have to see how it turns out.
I (honestly) don't want this to be just about me...
for now, I have appointed myself King of the Brain, but I don't need or want to do this on my own, so if you'd like to contribute to the error bar, then get in touch - email me at email@example.com, or tweet @BarError to pitch your idea and start recording. there's more on the website about what this might involve, and I encourage anyone, from GCSE science students to emeritus professors, to pitch a brain science story at me.
the format of the error bar will be all in 3's: episodes will be 9, 18, or 27 minutes long. 9 minutes of News followed by 9 or 18 minutes of discussion; each of the - 9 minute section will be made of 1 and 3-minute long pieces.
if you just want a second opinion, or to contribute a topic, you can email firstname.lastname@example.org and we can try to fact-check a brain story you've seen in the news.
on the name
the error bar is called the error bar partly because these sorts of conversations have started for me in pubs up and down the world over many years.
but an error bar is also a thing used a lot in statistics and science. an error bar is a margin of error, or a region or a range of uncertainty, that is almost always useful to have when you're trying to interpret a number.
on my favourite podcast - More or Less with Tim Harford - the host often asks of a statistic: 'Is that a big number?' That's a very important question. Whatever statistic or number is given, we need to know if it's big, if it's higher or lower than expected, if it means we have to make some changes based on this number, or if we can just ignore it.
instead of just talking about the size of numbers, the error bar will ask ' about all numbers ' what's the error bar on that? So, if you say that an average of 90% of middle-aged men across the world want to start a podcast, I'd want an error bar on that - give me the level of uncertainty you have about that number. If you asked every middle-aged man in the world, then you could say it was exactly 90%. But if you only asked a small number of men from, say, three or four different countries, then it's very unlikely that 90% is exactly the right number. you wouldn't get the same result if you did the survey again, or in a different country. so what's the range? Is it 70-95% of men across the four countries sampled? or what's the 'confidence interval' - a technical term for a similar thing - around that average number'
in summary: the error bar is where neuroscience meets uncertainty, with lounge music and Big Ben.
tune in next time for the news at brain
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 the University of Nottingham. 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.ε