The ancient marriage between music, movement and mood

Think back to that moment when you first heard your favorite song. What about it made you stop in your tracks? Was it the incessant buildup, soaring high, filling you with a sense of elation? The flirty high notes, light as wings, bringing a bounce in your step? Or the rumbling base drop, furiously cascading, sending shiver after shiver down your spine?


Feel the music. Source:

Music has always had a special place in my heart. Like many, I use it as an emotional outlet and a painkiller for physical aches. During one of my longer runs a while back I distinctively remember a sense of elation as a drum & bass piece with the perfect BPM came on. I matched my foot strikes to the beat, closed my eyes and ran in a state of pure euphoria. (It helped that I was one of the only people on a flat, spacious wall that hugged the ocean. I must’ve looked high.)

Since that run I’ve been musing over our relationship with music. Across the globe people describe intense pleasure from listening to music, grooving to music, exercising to music. What lies at the core of this abstract euphoria? What is it about our perception that allows us to experience all three in unison?

I hope to answer these questions with my new piece up at Scientific American MIND. It is science writing based on peer-reviewed literature; but it’s also my personal ode to music.

If you have a moment please check it out, and let me know what you think!

EditBeau Sievers, one of the authors of the study, kindly provided feedback on Twitter. She pointed out “The SciAm piece could be misread as saying Kreung music has no tuning, timbre, or scales—they do, they are just not Western.  Kreung mem music does have clear notes in it, articulated to evoke insect sounds—very ‘buzzy’ but still musical. Thanks for the very nice writeup!”

Hope this helped to clarify things!

Kudos to Virginia Hughes at National Geographic blogs (Phenomena: Only Human) for directing me to the cited studies. She has previously written about them individually on her blog. Her writing is FANTASTIC – if you’re not a reader yet, I highly recommend her work.


#SfN13 Running boosts cognition by increasing aerobic capacity

Poster: 671.Learning and Memory: Genes, Signalling and Neurogenesis II.


Running for health. Source:

There’s no doubt that aerobic exercise benefits the brain. Running, for example, reduces anxiety, improves sleep quality, boosts learning of a new task and maintains spatial memory*. Many of these mental perks stem from an increase in adult neurogenesis; that is, the birth of new neurons in the hippocampus and the olfactory bulb. (*That is, if rats run before new learning. See here for more.)

Yet perhaps the most apparent health benefit of running is increased cardiovascular and lung function. As any runner can attest to, an initially exhausting 10k soon becomes a breeze – you’ve increased your aerobic capacity. This led researchers from Duke University to wonder: is improving exercise capacity –by whatever means – necessary and sufficient to boost neurocognitive function?

Better bodies, better minds

Just like us humans, rats have an innate sensitivity to the effects of exercise. After the same 8-week running regime, high-response rats drastically increased their maximal capable running distance (~75%), while low-response rats barely improved (~22%). Surprisingly, compared to their sedentary peers, only high-response rats showed elevated neurogenesis in the dendate gyrus, a subregion of the hippocampus, as compared to their sedentary peers.

One hypothesized function of the dentate gyrus is pattern separation, or VERY simply put the discrimination between two very similar spatial contexts or things (Jason Snyder of Functional Neurogenesis fame has a great blog post on the matter). Researchers decided to challenge these rats with two Lego pyramids that only differed in the colour of their tops – imagine two Christmas trees with either a yellow or orange star. After the rats familiarized themselves with the yellow-topped Lego, researchers waited a minute before presenting them with both. High-response runners (but not their sedentary controls) instantly realized something was up – they approached and sniffed the new construct in earnest, ignoring the old familiar one.

Low-response runners, on the other hand, behaved just like their sitting peers, spending a similar amount of time with both objects. Low-responders had no problem with their memory; when faced with a mug and a can, they could easily discriminate between the two. They just couldn’t pick out minute differences in the Lego pieces, a skill often attributed to enhanced neurogenesis.

These data, perhaps somewhat dishearteningly, suggest that running doesn’t always boost brainpower – neurocognitive benefits only occur in tandem with improvements in aerobic fitness, as measured by total running distance until exhaustion. These results parallel that of a human study, in which increased lung capacity after training correlated with better performance on a modified pattern separation task (although understandably they did not show enhanced adult neurogenesis, so it’s hard to attribute behavioural output to increased new neurons per se).

Running-improved aerobic capacity seems to be the crux to exercise-induced brain benefits. But is running really needed? To explore this idea further, researchers decided to take treadmills out of the equation and focus on genetic differences in aerobic fitness.

Innate aerobic capacity accounts for cognitive benefits


Rats on treadmills. Source:

Allow me to introduce to you the low and high capacity runners. Selectively bred for their capability (or not) to “go the distance”, these rats differ up to 3 times in a long-distance standard fitness test, without ever setting foot on a treadmill. At 10 months old, they also had a two-fold difference in the total number of newborn neurons in the dentate gyrus as a result of increased neuron survival, which increased to three-fold at 18 months old.

Researchers took sedentary rats from both groups and challenged them to the Lego task described above. High capacity runners significantly outperformed their low capacity peers, expertly telling apart the Lego constructs. Similarly, in an object placement task in which researchers minutely moved one of two objects, low capacity runners could not identify the moved one after an hour’s delay, though they managed if the wait was only a minute. High capacity runners, on the other hand, excelled in both cases.

These results argue that high aerobic capacity in and of itself promotes pattern separation. But what if, unbeknownst to researchers, high capacity runners were maniacally jumping around everyday in their home cages? A few days of stealthy observation proved this wrong; paradoxically, low – compared to high- capacity runners were much more hyperactive. They also seemed more outgoing in a social interaction test, and exhibited a lower tendency to generalize trained fear from one context to another.

Running-induced neurogenesis is generally considered to ease anxiety. So why do high capacity runners (with higher rates of neurogenesis) seem more neurotic?

Born to laze, born to run


Sitting on a couch is really not that stressful. Don’t make me run!

Running is physiologically stressful in that it increases the level of corticosterone (CORT), a stress-response hormone. Unlike chronic stress that continuously elevates CORT, running only induces a transient, benign increase that quickly returns to baseline after recovery.

Researchers trained low- and high- capacity rats on treadmill running 5 days a week for a month. By the end, both groups showed increased running capacity, though trained low-capacity rats were only as good as untrained high-capacity ones (life’s unfair!). However, their acute stress responses drastically differed in a running-stress test.

Untrained low-capacity rats remained calm throughout the test, as measured by unchanging CORT levels. “They waddled on the treadmill for a bit, got tired and gave up.” said the researcher, “so they really weren’t that stressed out.” Trained low-capacity rats however hated the treadmill – their CORT shot through the roof. “You’re chronically forcing them to do something they’re terrible at, of course they’re going to be stressed out” explained the researcher, “and once they’re done, their CORT goes back to normal.” (I’m paraphrasing.) While this scenario is certainly possible, an alternative explanation is that only trained low-capacity rats were able to exercise to the point to induce a normal elevation in CORT levels; untrained rats simply don’t workout hard enough.

Intriguingly, untrained high-capacity rats had elevated levels of CORT during the running test, while previous training eliminated this response. Why? Researchers believe that chronic running habituated them to the stressor: “You know when you have this itch to run? You get stressed out when you can’t, and feel relieved when you finally do exercise.” In other words, these rats were “born to run”.

On the cellular level, running did not significantly increase neurogenesis in the ventral hippocampus in either low- or high-capacity rats, which I find rather surprising. Finally, high-capacity rats (compared to low) had less Mmneralocorticoid receptor (MR) and glucocorticoid receptor (GR) in the amygdala and hypothalamus, but not in the hippocampus. This is also surprising, as MRs and GRs in the hippocampus are crucial for negative feedback to the stress response axis (below).

Screen Shot 2013-11-16 at 7.50.14 PM

Hippocampal GR negatively regulates the stress response. Source:

Taken together, these data point to increased aerobic fitness– through genetic means or exercise- as the key to enhancing neurocognitive function in rats. Inbred differences in aerobic fitness may alter how one responds to exercise (and perhaps other types of) stress.

These studies beg the question: what if we could artificially mimic the effects of exercise (pharmaceutically or otherwise) and reap its benefits? While “exercise pills” may not necessarily benefit healthy individuals, they could potentially improve both physical and hippocampal health of the elderly or the disabled.

Such research is under way, though as of now the results are not yet convincing.

PS. This is the end of #SfN13 blogging. It’s been hectic, a bit overwhelming and a LOT of fun!! Thank you to all the presenters for your patience & feedback and the PIs who let me write about your work. Thank YOU for reading!

Regular research blogging will resume soon. Stay tuned!

671.01. CL Williams et al. Rats selectively bred for high running capacity have elevated hippocampal neurogenesis that is accompanied by enhanced pattern separation ability. 

671.02. KM Andrejko et al. Rats selectively bred for high running capacity have elevated hippocampal neurogenesis that is accompanied by a greater expression of hippocampal glucocorticoid receptors and altered contextual fear conditioning. 

671.04. JM Saikia et al. Treadmill exercise training only enhances neurocognitive function if it is accompanied by significantly increases in aerobic capacity. Duke Univ., Durham, NC; Univ. of Michigan Med. Ctr., Ann Arbor, MI

Exercise in a pill…again?

photo (14)

This is the view I run to 🙂  Stanley Park, Vancouver BC, Canada.

When life gets rough, I go out for a run. A sweet tune, a steady pace, a great view and all the anxiety and stress that plague me melt away. Though tough to get into, exercise is not only beneficial for your cardiovascular health, but also improves cognition and mood. For those who’ve suffered strokes, exercise may be even more effective than drugs in prolonging life.

Scientists have chased down the brain health benefits of exercise to a protein called brain-derived neurotrophic factor (BDNF). This protein nourishes existing neurons and promotes their survival; it also encourages the growth of new neurons and the formation of synapses, which neurons use to communicate with each other. In the hippocampus, a brain region vital for learning, memory and mood, BDNF supports the formation of long-term memories and is a crucial component of the molecular machinery that mediate anti-depressant effects.

Researchers have long searched for ways to increase BDNF without resorting to cardio, that is, to recreate the benefits of exercise in a pill. While this would not necessarily help the general public, it may allow those unable to engage in physical activity to enjoy the health of endurance without exertion. Unfortunately, BDNF itself is too large to cross the blood-brain barrier; it needs to be directly injected into the brain to work. Worth it? Probably not!

An alternative strategy is to focus on messengers that link the body to the brain. In other words, how does exercise, which happens in the body, increase BDNF in the brain?

Christiane Wrann et al (2013). Exercise Induces Hippocampal BDNF through a PGC-1α/FNDC5 Pathway. Cell metabolism. doi: 10.1016/j.cmet.2013.09.008

Enter our two candidate proteins, both of which are expressed in skeletal muscle and increase with exercise. PGC-1alpha prepares your muscles for longer and harder abuse by increasing the number of cellular energy factories (mitochondria), allowing better energetics. This protein is a transcription coactivator; that is, it promotes the expression of other downstream proteins, one of them being FNDC5, aka irisin. In the periphery, irisin “transforms” beige fat into the more metabolically active brown fat to support thermogenesis and maintain a healthy metabolism; in the brain, it seems to help neurons mature – but not much else is known. Could these two exercise-induced proteins be the messengers?

The researchers gave a bunch of male mice a running wheel, and watched them for 30 days. Lab mice are as enthusiastic about running as your average pet hamster; give them a wheel and they’ll go at it for hours. After their final bouts of exercise, researchers looked into their brains and found an increase in both proteins, specifically in the hippocampus (graph B below; green- running; black-sedentary) but not in the remainder of the brain (graph C). This boost of irisin paralleled a small increase in BDNF, again only in the hippocampus.

Screen Shot 2013-10-14 at 3.18.14 PM

Fndc5 is irisin; Erra is also an exercise-related protein which works with PGF-1alpha to increase gene expression.

Is irisin CAUSING more BDNF expression? Researchers used a virus to deliver a strongly expressed version of the irisin gene into isolated cortical (not hippocampal!) neurons in culture; this forces neurons to produce more irisin than they would naturally. As you can see below, this treatment (blue) caused an increase in BDNF, as well as MANY other proteins induced in neuronal activation (viral-delivered Green Fluorescent Protein/GFP, black bars, was used as a control).

Screen Shot 2013-10-14 at 5.05.43 PM

On the other hand, when researchers used three different short “hairpin”-like strand of synthetic RNA (shRNA) to inhibit irisin RNA expression, BDNF levels tanked (crimson bars) compared to the control (black bar).

Screen Shot 2013-10-14 at 5.07.16 PM

Since BDNF is linked to neuronal survival, researchers next checked to see how they faired. As predicted, forced expression of irisin increased the number of surviving neurons, while eliminating the protein had the opposite affect (though this could be due to toxicity of the shRNA). Further investigation found that BDNF inhibited irisin expression in a negative feedback loop, presumably to keep protein levels steady.

Neurons in culture can’t give us the whole picture. Researchers next turned to mice. When they forced the liver (why not skeletal muscle?) to overexpress irisin, once again BDNF increased in the hippocampus but not other areas of the brain. The effect size however, as you can see below, is TINY. Again, many other proteins were upregulated.

Screen Shot 2013-10-14 at 4.05.37 PM

So here we have a problem: irisin increases BDNF in the hippocampus in mice; nevertheless, researchers used cortical neurons for most of their studies. What happens if they switched to cultured hippocampal neurons instead?

The curse of negative data. As you can see below in graph D, stimulation of hippocampal neurons with synthetic irisin (blue bars) did NOT significantly raise BDNF levels (no star on the BDNF blue bar), though it did induce expression of many other proteins. When researchers eliminated irisin with three different shRNAs (graph E, Fndc5=irin, black bar is control; see how shRNAs cause a drop in its levels?), two of them also caused a decrease in BDNF as expected (shFndc5-1 and shFndc 5-2). The third? Not so much.

Screen Shot 2013-10-14 at 4.10.33 PM

So what’s the verdict? While exercise no doubt increases BDNF in the hippocampus, irisin alone most likely doesn’t. The data clearly shows that the effect size is either VERY small, as in their animal model, or non-existent! Furthermore, irisin is quite promiscuous: it increases the expression of other genes as well. These genes, known together as “immediate-early genes”, are expressed right after neuronal activation, and can also lead to an increase in BDNF levels. Hence it’s impossible to conclude that exercise increases irisin in the periphery, which in turn boosts BNDF in the hippocampus and bam! Less anxiety and better memory.

Irisin may mediate the benefits of exercise in skeletal tissue (this is also debated); but I wouldn’t bet on it as a potential exercise pill for brain health. Apparently, some people do – the lead author is chair of Ember therapeutics, a company that among other things focuses on irisin for the treatment of metabolic syndrome.

Scientists have so far uncovered several target proteins that recapitulate the effects of exercise when increased individually. Whether a single drug can mimic the mind-boggling spectrum of physiological and psychological effects of exercise though is another question. While scientists continue on their search, me? I’m going for a run.
Christiane D. Wrann, James P. White, John Salogiannnis, Dina Laznik-Bogoslavsk, Jun Wu, Di Ma, Jiandie D. Lin, Michael E. Greenberg, & Bruce M. Spiegelman (2013). Exercise Induces Hippocampal BDNF through a PGC-1a/FNDC5 Pathway Cell Metabolism : met.2013.09.008

Feeling anxious? Run it out!

Woman running by the ocean beach at sunset

Run away from stress! Source:

When life isn’t going well, I go for a run. I’ve always found running soothing. Maybe it’s due to “runner’s high”, the burst of endorphins that dampen physical pain and elevates mood. Maybe it’s because running increases the generation of new neurons in the brain (of mice), which we think is protective against depression.

Or maybe, as this new study shows, it’s because running tweaks the brain’s inhibitory circuits to directly dampen anxiety.

Schoenfeld et al. (2013) Physical Exercise Prevents Stress-Induced Activation of Granule Neurons and Enhances Local Inhibitory Mechanisms in the Dentate Gyrus” J. Neurosci. 33(18):7770-7777

Let’s first zoom in on the ventral hippocampus deep within the brain. This is one of the areas that process emotions, and is implicated in stress and anxiety regulation*. Increased activity in the ventral hippo is correlated (but not causative of) with anxious behaviour. Since running decreases anxiety, researchers wanted to know if runners’ ventral hippo respond differently to stress than sedentary people, in such a way that dampens anxiety.

(* You might remember the hippocampus is important for learning and memory – you’re right! However, increasing evidence is pointing to the dorsal hippo as the processing power behind memory. The hippo is quite a multitasker!)


Elevated plus maze. I even find it scary! Source:

Since directly monitoring brain activity at the single neuron level from people is impossible, scientists turned to mice. If you ever had a pet hamster, you know that rodents love to run – give them a wheel and they’ll go at it for hours. After 6 weeks of voluntary running, scientists placed these mice onto an elevated maze with two dark closed arms and two light open arms (imaging a cross-like mountain with cliffs at the ends, pic left). Runners showed significantly less anxiety as they explored the open “cliffs” than their sedentary peers. They also had more newly born neurons in their brains.

So running decreases anxiety, but is it through lowering hippocampus activation? To tackle this question, scientists exposed the mice to cold water. If you’ve ever tried a New Years polar bear swim, you’ll know that swimming in cold water is very stressful. Indeed, in sedentary mice, cold-water stress caused a spike in neuronal activity in the ventral hippo, as measured by a set of genes that transiently and rapidly get turned on in response to neuron activation. These immediate-early genes act as messengers to tell the neuron to start making proteins to adapt to the stimulus, and are a reliable sign of recent neuron activity.

As you can see below, couch-potato mice showed a spike in neuronal activity (Sed, black bar with a star), as measured by immediate-early genes c-fos and arc. This response was almost completely wiped out in the runners (Run, black bar with no stars). So running decreases ventral hippo’s willingness to react to stress, leading to less anxious behaviors. But how?

Screen Shot 2013-07-10 at 2.50.58 PM

The activity of neuronal circuits is mainly balanced by two antagonistic neurotransmitters: glutamate-mediated excitation and GABA-mediated inhibition. Most anti-anxiety meds right now work by increasing GABA signaling. Researchers found that runners had significantly more GABA neuron activation when exposed to cold-water stress. These mice also released more GABA neurotransmitter, especially during the period of stress (see the peak in the black line below?). So maybe increased GABA in runners is enough to increase inhibition and dampen ventral hippo activity?

One way to test this is to block GABA signaling and see how these runner mice behave. To test for anxiety, researchers brought back the elevated plus maze. As you may remember, this maze has two dark, chill closed arms, and two brightly lit open arms. Usually mice prefer to spend more time in the closed “safe” arms, and this is indeed the case with sedentary mice (white bar). However, runners showed increased exploration of the open “cliff” arms of the elevated maze just like before (black bar). They were way less anxious about the light and openness of those cliff-like arms.

Screen Shot 2013-07-10 at 2.55.09 PMNow, if you block GABA signaling with a chemical called bicuculine in runners, these mice (grey bar above) behaved just like sedentary mice (white bar). Their anxiety returned! Bicuculine only worked when given to the ventral hippo; if you block GABA in the dorsal hippocampus – important in learning and memory but not mood – it didn’t affect the runners’ anxiety levels. These results tell us that increased GABA signaling lowers ventral hippo activation, and this leads to decreased anxiety.

Overall the researchers pretty convincingly show that running reduces anxiety through activating GABA signalizing in the ventral hippocampus. It would’ve been nice to see how runner vs sedentary mice behaved in the maze AFTER cold-water exposure, ie if running can “immunize” mice against stress-induced anxiety as well. It would also be interesting to see how long this anti-anxiety change lasts once the runners stopped running – does GABA signaling go back or does it stay responsive for a long time?

Regardless, this study gives you another reason to go out for a run and keep running. Try it for three weeks (how much the mice ran) and see if it helps with stress and anxiety. Science says it does.
Schoenfeld TJ, Rada P, Pieruzzini PR, Hsueh B, & Gould E (2013). Physical exercise prevents stress-induced activation of granule neurons and enhances local inhibitory mechanisms in the dentate gyrus. The Journal of neuroscience : the official journal of the Society for Neuroscience, 33 (18), 7770-7 PMID: 23637169