Uncoupling of brain glucose uptake during walking in MS


Walking speed and brain glucose uptake are uncoupled in patients with multiple sclerosis

John H. Kindred, Jetro J. Tuulari, Marco Bucci, Kari K. Kalliokoski and Thorsten Rudroff
Front. Hum. Neurosci., 18 February 2015 | doi: 10.3389/fnhum.2015.00084


Motor impairments of the upper and lower extremities are common symptoms
of multiple sclerosis (MS). While some peripheral effects like muscle
weakness and loss of balance have been shown to influence these
symptoms, central nervous system activity has not been fully elucidated.
The purpose of this study was to determine if alterations in glucose
uptake were associated with motor impairments in patients with multiple
sclerosis. Eight patients with multiple sclerosis (four men) and eight
sex matched healthy controls performed 15 min of treadmill walking at a
self-selected pace, during which ≈322 MBq of the positron emission
tomography (PET) glucose analog [18F]-fluorodeoxyglucose
(FDG) was injected. Immediately after the cessation of walking,
participants underwent PET imaging. Patients with MS had lower FDG
uptake in ≈40% of the brain compared to the healthy controls (pFWE-corr < 0.001, qFDR-corr < 0.001, ke = 93851) and walked at a slower speed [MS, 1.1 (0.2), controls 1.4 (0.1), m/s, P
= 0.014]. Within the area of lower FDG uptake 15 regions were
identified. Of these 15 regions, 13 were found to have strong to
moderate correlations to walking speed within the healthy controls (r > −0.75, P < 0.032). Within patients with MS only 3 of the 15 regions showed significant correlations: insula (r = −0.74, P = 0.036), hippocampus (r = −0.72, P = 0.045), and calcarine sulcus (r = −0.77, P
= 0.026). This data suggest that walking impairments in patients with
MS may be due to network wide alterations in glucose metabolism.
Understanding how brain activity and metabolism are altered in patients
with MS may allow for better measures of disability and disease status
within this clinical population.

I normally consider PET studies to be an expensive way of finding out what we already know…but if its a sledgehammer that you need to get a point across, then who am I to question it?!  

In essence fMRI (functional magnetic resonance imaging) is used to find out what the brain is doing when you perform a motor task using radiolabelled glucose ( [18F]-fluorodeoxyglucose (FDG)) that is injected during the activity intravenously (i.v.). The FDG uptake into the brain then measures the brain activity. 
Not surprisingly in this study the self-selected walking speed of MS participants on the treadmill was slower than normal controls, so they made the some of the controls to walk slowly as well to avoid walking speed being a confounding factor in the study. 

Overall, MSers showed reduced FDG uptake into ~40% of the brain. Only three (insula, hippocampus and calcarine sulcus) out of 15 (others being frontal cortex, occipital cortex, motor cortex, lateral temporal cortex, medial temporal cortex, cerebellum, anterior cingulum, precuneus, lingual, fusiform, thalamus and caudate), compared to 14 out of 15 regions in controls, were found to be correlated with waking speed (see figure below). Therefore, there is a decoupling of brain glucose utilization and motor task performance, i.e. the brain network connections are altered in MS. 

It is important to realize that x does not equal y as far as brain activity is concerned, brain activity is a huge network of nodes and connections (similar to those air route maps you see in the back of in-flight magazines!); functionality is therefore very much dependent on network efficiency. If network efficiency is affected in MS brains, can exercise encourage neuroplasticity and improve on this? Maybe this is an exercise that still needs to be done?!

Figure: Visual representation of association between walking speed and brain region FDG uptake (MS=filled in dots, controls=open dots). In each instance the strength of association was less in MSers (bottom r value) than in controls (top r value). r is a coefficient measure of strength of association with a value of 1 being a direct association. SUV=standardized uptake values.

If other work along the same line interests you, please also check out my post on the 24th Feb 2014 – ‘Is MS starving the brain?’.

About the author

Neuro Doc Gnanapavan


  • Has the decoupling of of brain glucose utilization during cognition been explored? It would be interesting to see if cognitive exercises also show a reduced neural network in MS.

  • Yes Steve, this has already been done. In MSers with memory problems there's reduced FDG uptake in the hippocampus, cingulate, thalamus, cerebellum which are areas of the brain associated with memory. Related tractography studies using MRI diffusion waited imaging (DTI) has demonstrated reduced network efficiency in the neural network related to the PASAT test (which is a serial addition test) irrespective of gender and estimated premobid IQ.

  • Does decrease glucose utilization track with hypoxia? I assume glucose consumption in the brain is not via glycolysis
    And a minor comment: in second paragraph under the figure, it's probably PET, not fMRI (why would one need radiolabeled glucose for MRI)

  • It's difficult to say, the study referring to hypoxia which I link above from a previous blog is not as sophisticated as this ones, and generally alludes to an overall reduction in FDG uptake. This work looks at regional specific utilization based on task and demonstrates nicely the disproportionality. I'm sure more will be published in the year to come on hypoxia in MS, I (and others!) believe its the hottest thing to come out in MS over the past two years.

    With regard to your glycolysis question, in fact the brain consumes around 20% of the body's energy supply, and is almost exclusively reliant on aerobic glycolysis, even in the resting brain. Interestingly, an increase in neuronal activity increases blood flow and glucose utilization to boost aerobic glycolysis even at the cost of oxygen consumption!! The brain has evolved through the years to disproportionally protect its energy source – which is why a stroke has devastating consequences.

    Apologies on the confusion about the two modalities, I did mean fMRI as I was referring to mapping of structural neural networks and cognition in MS, and I was trying extrapolate from this work!

  • Decoupling of brain glucose utilization with motor and cognitive impairments may be reflective of mitochondrial dysfunction (seen in neurodegenerative disease). Re-routing of neural pathways to compensate for this loss in energy metabolism would be expected in axonal injury whether inflammatory or traumatic. This blog has posted much on the role of mitochondrial dysfunction and decreased neural potentiation after demyelination. Can exercise increase neuroplasticity? I think it has been shown that neural networks can be created following mental "gymnastics". Use it or lose it.

    • Exactly, mitochondria may be the missing link. Boosting function with Coenzyme Q10 as we do in mitochondrial disorders may be one option. I think mouse doc has previously posted on Co Q10 ameliorating inflammation before. It's definitely cheap – a quick browse through Amazon, Co Q10 prices seem to be under £20.

    • Doesn't CoQ10 stimulate the immune system?
      Don't people with MS want to avoid doing that, just like they want to avoid infections?

  • There was work presented during last years ECTRIMS/ACTRIMS which demonstrated improved thalamic connectivity following Dr Kawashima's brain training! Professor Dawn Langdon seemed perplexed when I mentioned Sudoko at a recent talk – so maybe that doesn't work!!



Recent Posts

Recent Comments