How Good is Exercise for Parkinson's?


by Carl Sherman

July 18, 2012

Exercise is good for people with Parkinson’s disease. “There are probably two or three dozen papers and countless anecdotes documenting that people who exercise have a reduction in symptoms,” says Michael Zigmond, professor of neurology at University of Pittsburgh. A 2008 meta-analysis of clinical studies linked exercise to improved balance, walking speed, strength, and overall physical functioning and quality of life. There are indications that cognitive function benefits as well.

But some researchers and clinicians hope for more—much more:  They think exercise may alter the Parkinson’s disease process itself, to slow or stop  progressive deterioration that leads to severe disability and, in many cases, dementia. This is something that no drug or surgical procedure, as yet, can do.

A convincing clinical trial to confirm this hypothesis, which would be lengthy and costly, has never been done.  But the weight of indirect evidence is imposing, according to a review paper, “Does vigorous exercise have a neuroprotective effect on Parkinson disease?”published in  Neurology in 2011.

“You can come at this from a variety of perspectives,” says author Eric Ahlskog, professor of neurology at Mayo Clinic, in Minnesota. “If you look at all the animal literature and the observational studies [in humans]…. Everything lines up to suggest you do good things for the brain with exercise.”

How exercise works

In Zigmond’s lab, experimental work over the past decade, summarized in a paper in the January 2012 issue of Parkinsonism and Related Disorders, has shown that in rats, mice, and non-human primates exercised in various ways before and after injection of a neurotoxin into the substantia nigra (the standard animal model for PD), the loss of dopamine neurons is “dramatically reduced,”  compared with animals given the same injection without exercise. The exercised animals also suffer less motor impairment.

“We want to understand how exercise protects,” Zigmond says. “There are many possibilities, but we’re spending most of our time [investigating] trophic factors,” such as brain-derived neurotrophic factor (BDNF) and glial cell-derived neurotrophic factor (GDNF). Work at Zigmond’s lab and elsewhere has shown that these compounds, which promote new synapse formation and support neuron survival, increase with exercise in animal brains.

In humans, post-mortem studies have shown reduced BDNF in brain areas affected by PD. And evidence linking exercise to higher levels of the protein in the body suggests (while it doesn’t prove) that levels of BDNF increases in the brain as well. 

In an animal study now beginning, Zigmond and his colleagues will use molecular approaches to probe the connection between exercise and trophic factors. Knocking out the gene for a single protein is unlikely to eliminate the exercise effect altogether, but a reduction, he says, would be “a fairly convincing indication that exercise protects through trophic factors.”

 Zigmond cautions that the animal model he and others use provides but a rough approximation of human PD. “Parkinson’s is a very complex disease that involves progressive loss of many different nerve cells.” While the death of dopamine neurons is believed responsible for PD’s characteristic motor symptoms—tremor, slowness of movement, rigidity— non-motor manifestations that include cardiovascular abnormalities and, late in the disease, dementia  implicate other brain regions and  neurotransmitters.  

Could exercise protect these systems as well? Zigmond thinks it highly likely. Besides its apparent effect on growth factors, “exercise increases blood flow in the brain, and reduces inflammation and oxidative stress. In fact, pretty much everything it does is the opposite of what happens in PD.

“If and when we can mount a good clinical trial,” he says. “I’d be astonished if we weren’t able to show that exercise is effective in slowing the disease.”  

Back to basics

Giselle Petzinger, a  University of Southern California neuroscientist, isn’t so sure. “Of course, the vision at the end of the tunnel is making a big difference in the lives of people with PD. But to get there we have to take a couple of steps back to see what’s relevant to change.”

She sees exercise in a wider role: a tool to unravel the complexities of PD pathophysiology, which could lead to effective treatments including novel drugs. “Dopamine is just a player, not the fundamental problem. The fundamental problem is a circuit that’s breaking apart. Exercise gives us new set of eyes to look at PD on a systems level.”

In animal experiments in her lab, exercise didn’t preserve dopamine neurons, but increased dopamine receptors in areas regulating movement. “The dopaminergic system changed in a way to [become] more efficient. … a compensatory effect that one could use to regain meaningful function.” In a study just getting underway, PET imaging seems to show a similar receptor increase in PD patients who exercised, compared to those who did not.    

Other recent work in Petzinger’s lab found that exercise modulated receptors for a different neurotransmitter, glutamate. “The problem in PD is that circuits encoding information for automatic movements and skills are impaired, and what drives these circuits is the glutamate system.”

One consequence of abnormal glutamate neurotransmission, she says, may be hyperexcitability in movement-regulating  regions of the cortex, previously observed in PD and possibly involved in motor symptoms. In a small human study, she saw reductions in cortical excitability, after 8 weeks of exercise, along with improved symptoms.   

The best exercise?

Among the fundamental questions about PD awaiting resolution, Petzinger says, are pragmatic ones like the intensity and type of exercise  (skill building, aerobic, or both) most likely to be effective.

Jay Alberts, of the Cleveland Clinic Lerner Research Institute, thinks he may have found an answer on a tandem bicycle— serendipitously, while pedaling across Iowa with a PD patient “to demonstrate that … an active lifestyle can and should be maintained after diagnosis.”

As reported in his 2011 paper in Exercise and Sports Sciences Reviews, the patient’s symptoms  improved markedly, and Alberts wondered if the “forced” nature of the exercise—as second rider, or “stoker,” she pedaled at a rate that he, as “captain,” determined—was responsible.    

“In  forced exercise, the individual is assisted in achieving a higher rate of [motion] than they could normally achieve,” Alberts says. “It’s still aerobic—they’re not just passively moving their legs.”

A pilot study followed, in which 10 PD patients were randomized to a program of “voluntary exercise” (VE) at their own rate on a stationary bike, or forced exercise (FE) on a stationary tandem bike with a trainer. While the FE group pedaled faster than the VE patients, resistance was adjusted to maintain a comparable heart rate. The groups reported similar levels of exertion.   

After 8 weeks, the FE group showed significant improvements in PD symptoms like slowness of movement, while the VE group did not. Although the exercise involved only the lower body, benefits included manual dexterity, Alberts says. “This suggested that something central was going on in the brain, and led us to do an fMRI study.”  

When unmedicated patients were examined shortly after a forced exercise session, their brain activation patterns resembled those seen when they were on medication. “People like to say that ‘exercise is medicine,’ and it looks like we’re actually showing it,” he says. 

Alberts is now recruiting  patients for an expanded trial that will test cognitive function along with motor symptoms after eight weeks of  FE, VE, or no exercise, and then eight weeks later. While this kind of study will not resolve the question of long-term, disease-modifying effects of exercise, it may suggest some practical answers.

“People often ask me: ‘what’s the best exercise for PD?’ At the moment, my answer is ‘any exercise that you’ll do,’” Alberts says. “Hopefully, we’ll soon have more data to give a very specific recommendation.”