Welcome to this week’s Sunday Science stories, with a new look at diabetes, novel approaches to brain injury and Alzheimer’s, brainy birds, and more…
Traditionally, diabetes has been classified as type I – usually early onset, autoimmune destruction of the insulin-secreting cells – or type II – usually late onset, and due to tissues becoming insulin-insensitive (insulin resistance), strongly associated with obesity. A paper in Lancet Diabetes and Endocrinology analysing data for adults with newly diagnosed diabetes is making the case that there could in fact be several distinct subtypes of “Type II” diabetes (note this is paywalled – an accessible news article can be found here). The subgroups responded differently to different medications, and had different risk factors for complications such as kidney disease. I think this is important, but needs refining and examining over other cohorts to see how robust it is – I think there’s a risk that simply putting people in smaller diagnosis boxes won’t necessarily help people who don’t fit neatly into one box, or who might move from one category to another. Diabetes is a complicated disease and ideally you want a proper role for personalised medicine so that individuals can have tailored approaches for this long-term condition.
“Bird-brained” is used as an insult, but birds may be smarter than we think. There is already a lot of research showing that corvids (crow family, pictured) are very smart; now it’s been shown that birds and mammals both have some brain cell types linked to intelligence. Science Daily version here.
The US Food and Drug Administration (FDA) has approved the first blood test for detecting concussion, of particular concern in the US with regards to American football players, who are at high risk of developing chronic traumatic encephalopathy (CTE) as a result of repeated head injuries. How useful this will be for CTE I’m not sure – it may be of more use in the A&E (ER) in emergency admissions. Meanwhile, research into being able to effectively distinguish CTE from conditions with similar symptoms is ongoing, particularly as both seem to involve raised levels of altered tau protein, implicated in Alzheimer’s.
Speaking of Alzheimer’s, and something that really sounds like it came out of a science fiction B movie….can flashing lights help treat neurodegenerative conditions by altering brainwaves? Yes, really – at least with some early promising results in mice, and some early trials in humans expecting results this year. Essentially, it’s known that conditions such as Alzheimer’s and Parkinson’s lead to altered patterns of brainwave activity, and that oscillations of light at certain frequencies can stimulate certain brainwaves. The beauty of this approach is that it’s completely non-invasive. A fascinating review of the field of neuromodulation to be found here.
A good review of deep learning AI systems being applied to analysing the mountains of biological image data that are being generated and need analysing. I somewhat over-confidently predicted to some of my students that a live cell-tracking programme that some colleagues had developed to analyse my data was the old, difficult way and AI was the future – I still think this is so, particularly as in many fields there’s getting to be more data than humans can handle, but it’s not all easy.
Work on how rocks draw carbon out of the air shows how current “negative-emissions strategies” remain merely dreams at the moment, making the importance of decarbonising human societies to avoid catastrophic climate change even more important.
And finally, a nice obituary of the writer Ursula Le Guin can be found in Nature magazine here, focussing on the anthropological side of her stories and upbringing. I’ve frequently noticed an appreciation for the ideas of the great science fiction writers amongst scientists: the inspiration goes both ways.