This week we look at a new study of blood proteins that may give rise to tests for early dementia diagnosis. We discover how a range of changes by the US EPA has resulted in a rapid rise in the amount of a toxic pesticide in the blood of Americans. We investigate a new computer chip that uses light rather than electricity to compute AI functions. Finally we examine the findings of a COVID-19 study which provided a surprising insight into how poor internal regulation of our body temperature is correlated to depression.
Early Dementia Diagnosis
We still do not have effective treatments for dementia. Many researchers believe this is due to the late diagnosis of all cases. An earlier diagnosis prior to the disease having taken hold may allow novel treatments to slow or potentially stop progression of the disease. The WHO estimates that there are currently 55 million people living with dementia, a number that will grow significantly in coming years.
A research team in the UK have analyzed blood samples from 50,000 healthy adults (via the UK Biobank). 1,417 of those adults later developed dementia over the 14 years post sample taking. An analysis of 1,500 blood proteins identified 4 that were strongly associated with dementia and Alzheimer's disease. For some participants the protein levels were outside normal ranges more than 10 years before the symptoms of the disease were identified.
One of the four proteins, GFAP, has already been identified as a biomarker of developing Alzheimer's. People with outside normal levels of this protein are twice as likely to develop the disease.
The team developed a machine learning algorithm that combined the levels of the four biomarkers with age, sex, education level and family history. The model was trained on 2/3rds of the samples available and then tested on the other 1/3rd. The model predicted the onset of three subtypes of dementia with 90% accuracy by using samples more than ten years before participants were officially diagnosed.
The team hopes to develop simple blood tests that will allow doctors to screen for disease risk but also to differentiate between subtypes of dementia. This will allow early intervention and potentially earlier treatments.
Toxic Chemical in the Blood
A research team at the Environmental Working Group (EWG) in the US has tested the urine of 96 people for chlormequat, a toxic pesticide. This chemical is known to cause reproductive problems in animals and it is thought that it likely to have a similar effect on humans.
The urine samples were collected between 2017 and 2023. The tests found that the amount of chlormequat was in more people and at higher concentrations as the years progressed. This suggests that consumer exposure to chlormequat is increasing. The EPA in the US changed the law in 2018 to allow chlormequat on imported oats (most imported oats in the US come from Canada). The amount allowed was increased again in 2020.
In 2019, Chemical manufacturer, Taminco applied for approval to use chlormequat on barley, oats, critical and wheat grown in the US. In 2023 the EPA proposed a rule for this to be approved. It has not yet been finalized.
The EWG tested oat based products in 2022 and 2023 and found numerous non-organic oat based products contained chlormequat. Organic oat products had little to no detection of chlormequat. More needs to be done to determine the impact on human health and the safe levels of the use of the pesticide.
Fitness trackers find a new symptom of Depression
A group of researchers at UC San Francisco commenced a COVID-19 research study in 2020. 20,880 people from 106 countries signed up to participate in the TemPredict Study. The team was trying to determine if fitness and health trackers could be used to detect early symptoms of COVID.
All participants wore an Oura Ring that measures heart rate, body temperature, oxygen levels and various fitness indicators. The participants were followed for 7 months. Data collected by the ring was stored in a phone app and the participants self reported their emotions and mental health in daily surveys.
The team was able to predict COVID infection 2.75 days prior to participants testing positive. The other finding that the team was not expecting, was that participants with higher body temperatures whilst awake, reported higher rates of depressive symptoms and feelings of depression. A previous study had found that the awake-asleep body temperature difference was more than twice as large in individuals in a control group verses a group with depression. This may indicate that poor body temperature regulation is somehow associated with depression.
Our body temperature naturally varies over time and throughout the day. People with depression find it harder to self cool their bodies. This may lead to non-pharmaceutical treatments. Ice baths can cool the body down as can heating the body up. Once heated, the body will naturally rebound and lower body temperature. This heating effect lasts longer than jumping in an ice bath.
Depressive symptoms are rising in society. If poor thermoregulation is a symptom then body temperature can be monitored and temperature treatments used when they would be most effective. If you are using an Oura ring (or similar product that measures body temperature) then a signal could be sent to you via a phone app indicating that it is time for you to have a hot bath or sauna.
Light Speed Chips
A team at the University of Pennsylvania have developed a new computer chip that uses light to perform the math required to train AI. The chip may also increase processing speed of standard computers and reduce their energy consumption.
Called a Silicon-photonic (SiPh) chip it brings together previous work that described how to manipulate materials at a nanoscale to perform math calculations using light. The team’s goal was to develop a platform for vector-matrix multiplication. this is used widely in neural networks that power many of our AI tools.
The team varied the height of the silicon wafers in specific regions on the chip. A 150 nanometer variation allowed the means to control the propagation of light through the chip. This variation in height can be distributed in order to cause light to scatter in specific patters. This allows the chip to perform calculations at the speed of light.
There are more advantages of the chip in addition to the speed of calculation. Many calculations can happen simultaneously removing the need to store information in the computer’s working memory. This lack of storage in working memory will make the technology virtually unhackable (hacking usually accesses working memory, however remember the weakest link in any computer system is the user, it doesn’t matter how impenetrable the system is if you let the hacker in the front door, watch what you click on).
The new chip may also be suitable for use in graphics processing units (GPUs) in our computers. This will help with the use of the many AI tools that are emerging.
Paying it Forward
If you have a start-up or know of a start-up that has a product ready for market please let me know. I would be happy to have a look and feature the startup in this newsletter. Also if any startups need introductions please get in touch and I will help where I can.
If you have any questions or comments please comment below.
I would also appreciate it if you could forward this newsletter to anyone that you think might be interested.
Till next week.