menu Contact Us Dizzy Patients Health Care Providers Dizzy Culture Research Site DVD BPPV DVD Tai Chi DVD Understanding Dizziness Acknowledgements Disclaimer Quoting

Progressive Supranuclear Palsy

Timothy C. Hain, MD Page last modified: July 20, 2013

Progressive supranuclear palsy (PSP) is a degenerative neurological disorder of uncertain etiology characterized by gait ataxia, slowing or inability to generate voluntary saccadic eye movements, and axial rigidity. The most characteristic aspect of PSP is an inability to move the eyes, but the first symptom of PSP is usually unsteadiness and falling.

PSP was first described as a distinct disorder in 1964. It is sometimes referred to as Steele-Richardson-Olszewksi, or "SRO", syndrome, from the names of the individuals who defined the disorder.

Who Gets PSP?

PSP is estimated to affect about 4-6.4/100,000 persons, or about 5-6 percent of persons thought to have Parkinsonism. The incidence rate for new cases for ages 50-99 is 5.3/100,000, the crude incidence rate is 1.1/100K (Bower et al, 1997; Schrag et al, 1999). The peak incidence is in the early sixties. Men are affected slightly more often than women.  PSP does not generally run in families, although there are some pedigrees reported.

An illness resembling PSP is very common in Guam, sometimes in association with ALS (amyotrophic lateral sclerosis) and dementia. The cause of Guamanian PSP is unclear although it has been attributed to unusual dietary factors, possibly interaction with genetic factors (Cox and Sacks, 2002). Recently a high prevalence of PSP like illness has been found on the island of Guadeloupe in the French West Indies (Caparros-Lefebre, 1999). In this population, it is suspected that a PSP like illness is related to ingestion of native teas called "soursop" and "sweetsop", both of which are forms of the "custard apple".

A famous person with PSP was Dudley Moore, the actor.

What are the symptoms of PSP ?

The symptoms and natural history of PSP was recently reported by Nath and others (Nath, Ben-Shlomo et al. 2003). The most frequent first symptom of PSP is several falls over a year. Next patients often develop some stiffness and at this point may be diagnosed as having "atypical Parkinsonism. However, patients with PSP rarely develop the resting tremor and stooped posture characteristic of Parkinsonism.

As the disease progresses, most patients will develop problems controlling eye movement. Double vision is reported in more than half of all patients (Nath, Ben-Shlomo et al. 2003). The eye problems begin with vertical eye movements -- patients may be unable to look downward.  This may result in the so-called "dirty-tie" sign, because patients can't see that they are dropping food when they eat. Difficulty in reading is common. Photophobia is reported in 43% (Nath, Ben-Shlomo et al. 2003).   Eventually, patients lose the ability to look up and down at all, and usually about a year later, the ability to look from side-side is also lost. Eyelid apraxia occurs in 43% (Nath, Ben-Shlomo et al. 2003).

Typically, patient with PSP have trouble controlling the sitting down process -- they may "fall into their chair". There is recent evidence for abnormal otoligh responses -- small translational-VOR, and reduced VEMPs (Liao et al, 2008)

Swallowing difficulties are also common in PSP (Litvan, 1997), the most common problem being delayed initiation of swallowing. The course of PSP was recently studied (Santacruz et al, 1998).

There are a number of possible "signs" of PSP that will need confirmation. Ghika and Bogousslavski suggested that presymtomatic hypertension is a major feature in the diagnosis of PSP (1997).

What causes PSP?

It is known that the symptoms of PSP are caused by gradually progressive damage to a group of cells in a part of the brain called the "midbrain". These cells are involved in eye-movements and balance. The cause of the degeneration of these cells is unknown. In addition to the midbrain disease, there is also damage to the basal ganglia (especially globus pallidus), subthalamic nuclei, and the dentate nucleus of the cerebellum. According to Cordato et al (2000), atrophy of the basal ganglia is largely confined to the internal globus pallidus. Cerebral cortex is also affected and decreased metabolism of cerebral glucose correlates with dementia. Cortical benzodiazepine receptors are also decreased (Foster et al, 2000).

Pathologically, gross examination of the brain in PSP shows midbrain atrophy. There is neuronal loss and neurofibrillary tangles in the basal ganglia, diencephalon and brainstem. The substantia nigra, subthalamic nucleus and pontine base are typically involved as well as the ventral anterior and lateral thalamic nuclei. The cerebellar dentate nucleus may show degeneration. Cortical pathology is minimal except for motor areas. See the review in Jellinger (1992) for more detail.

Until recently, the main suspicion for cause fell upon either a virus or a slow toxin.  For example, a toxin called "MPTP", a contaminant in a drug of abuse, causes a condition similar to Parkinsonism. It has been speculated that there may be other slow-toxins in the environment, as for example, cycad nut or fruit bat consumption in Guam (Cox and Sacks, 2002) and certain herbal teas used in the Caribbean. With respect to the virus hypothesis, certain variants of Parkinsonism are known to be related to strains of influenza, and it is conceivable that a so-far undescribed virus is the cause of PSP.

Recent genetic studies, however, suggest that some cases of PSP is an autosomal recessive condition that maps to a polymorphism in the tau gene. (Bennet et al, 1998; Higgins et al, 1998-1999; Spillantini and Goedert, 2001). Tau is a microtubule-binding protein that is normally abundant in neurons. There are six different forms of tau in normal human brain. In typical PSP, pathological tau is composed of aggregated 4-repeat (E10+) forms that accumulate in cells and glia in the brain (Searceant et al, 1999; Spillantini et al, 1998). Rojo et al (1999) recently reported 12 pedigrees with familial PSP. Relatives of patients with PSP tend to score more abnormally on screening tests for Parkinsonism (Baker and Montgomery, 2001), supporting either a genetic factor or exposure to a common environmental toxin. Genomic screens in persons with late-onset Parkinsons disease also suggests a linkage a mutation on the Tau gene on chromosome 17q (Martin et al, 2001)

Delacourte et al suggested that tau is not the primary problem in most neural degenerations, but rather is a marker for vulnerable neurons that are damaged in several degenerative diseases (1998). Perhaps consistent with this line of logic, there are a multitude of other "tauopathies" including Alzheimer's disease, Picks disease, ALS-Parkinson dementia complex of Guam, familial "tauopathy" (Murrel et al, 1997) , and corticobasal degeneration (Higgins et al, 1999). There are four distinct kindreds with a mutation on chromosome 17 called N279K, and 50 kindreds having a syndrome called FTDP-17 for frontotemporal dementia and parkinsonism (Arima et al, 2000; Wszolek and Hutton, 2000), that lumps together 9 different mutations (Reed et al, 2000). The tau in PSP is different from that observed in Alzheimer's disease and Picks, both in morphology and tau isoform content, but it resembles the tau in corticobasal degeneration (Di Maria et al, 2000; Houlden et al, 2001) and FDTP-17. PSP certainly need not be entirely genetically determined -- it also seems possible that PSP is partially controlled by genetic susceptability and also partially related to other stressors such as toxins or viruses. Oxidative stress, perhaps related to mitochondrial disorders, is another possibility.

Slow downward saccades in a patient with PSP (unpublished data of Dr. Hain). These were recorded with a scleral eye coil.

Supplemental material on the site DVD: Video of slow vertical saccades in patient with PSP (10 meg)

A PSP-like disorder has been reported after surgical repair of the ascending aorta (Mokri et al, 2004). This condition closely resembles PSP, but appears within weeks to months following this type of surgery.

Slow downward saccades in a patient with a PSP like syndrome following aortic surgery(unpublished data of Dr. Hain). These were recorded with an infrared oculography system in our dizziness clinic in Chicago.

How is PSP diagnosed?

PSP is a clinical diagnosis, meaning that there is no absolute "test" for PSP. PSP is ordinarily diagnosed by a neurologist who has had experience with this condition. If anything, limitation or slowing of vertical saccades (see above) is the closest to a "litmus" test. Good recordings of vertical eye movements are not that easy, and to get this done properly requires good equipment

Recently it has been reported that a particular type of MRI scan, Diffusion-Weighting (DW), can differentiate PSP from Parkinsonism based on increased diffusion coefficients in the putamen. However, this type of MRI does not distinguish PSP from the Parkinson variant of multiple system atrophy (MSA). (Seppi, Schocke et al. 2003). In our opinion, while DW MRI should ideally be done as a confirmatory test when the PSP patient presents, there is little clinical usefulness in the DW scan as there are no effective treatments for PSP or MSA, and differentiation of PSP from Parkinsonism is not difficult for an experienced clinician. The superior cerebellar peduncle is also, on average, 20% smaller in PSP than in controls (Paviour et al, 2005). Practically this would seem to be a difficult distinction to make in clinical material.


The ratio of the area of the midbrain to the pons on saggital MRI has reported to be an accurate method of detecting PSP (Oba and others, 2005), as the midbrain is smaller in PSP. This method was also supported by Massey et al (2013), who suggested that a ratio of 0.52% was "100% specific" for PSP. Although this method is supposedly simple, the methodology of computing "elliptical regions of interest", seems to us to involve considerable subjective judgment on the part of the radiologist and also, the overlap between controls and PSP patients seems to us to be substantial. We suspect that the problem might be the intrinsic doubling in uncertainty involved when dividing one uncertain number by another. We also think it is imprudent to ask a radiology doctor who does not examine the patient, to determine from a somewhat subjective measurement, whether or not a patient has a deadly disease. We think that there would likely be an tendency to "undercall".

Another approach is a spinal fluid test developed in Italy for "truncated Tau forms" (Borroni et al, 2008). This spinal fluid test, not currently available at a clinical test, was reported as extremely sensitive and specific for PSP. As there is no commercial method of doing this test, it's clinical utility is presently absent.

There are several very obscure neurological diseases that can be confused with PSP (see table below), and it is wise to seek out a "tertiary care" neurologist who is familiar with PSP. This usually entails being referred to a "movement disorders clinic" in a university hospital setting. Experienced neurologists are generally accurate in making the diagnosis of PSP, being "right" about 90% of the time. General practitioners often misdiagnose PSP as Parkinsonism (Nath, Ben-Shlomo et al. 2003).

Eye movements, particularly saccades (see figure and movie above) are nearly always abnormal, but there are other (rare) causes of slowed eye movements. Classically it is taught that in PSP there is an "axial rigidity", meaning that the limbs may be relatively normal while the neck and trunk are rigid. This idea was called into question recently by a paper by Tanigawa that suggested that only the neck is rigid (Tanagawa et al, 1998). There are some unusual tests that may be positive, such as the "applause sign" -- see next section. Blood tests, CT and MRI scans are usually normal. 

Tilt table and Valsalva testing may be abnormal in PSP as well as related disorders such as MSA. (Schmidt et al, 2009).

A large number of other radiological techniques have been recently suggested (Golbe, 2004). Several of these depend on documenting subtle shrinkage of the midbrain and related areas using MRI (Paviour et al, 2005). Other recently reported methods include differences in the amplitudes of transcranial magnetic stimulation, and SPECT scan. At this writing (6/2005), we feel that these methods are of no practical utility and also need confirmation.

How is PSP different from Parkinson's disease?

All patients with PSP have abnormal vertical eye movements. Patients with PSP rarely have tremor and the stooped posture characteristic of Parkinson's disease. Another degenerative disease, Gaucher's type-III, a metabolic storage disese, also causes a progressive supranuclear palsy, but it begins horizontally. Other disorders that may be mistaken for PSP include corticobasal degeneration, Picks disease, multisystem atrophy (MSA) and diffuse Lewy body disease. There is also a recently described variant of myotonic dystrophy (MD2), which has some resemblence to PSP. MD2 affects tau (Maurage et al, 2005).

Links to pages about these disorders in on the table below. A familial "tauopathy" with dementia closely resembling PSP was recently described (Murrell et al, 1997). Jacob-Creutzfeld disease and related disorders can also present with ataxia and a supranuclear gaze palsy, but the course is generally more rapid and dementia is more severe than in PSP.

MRI may be able to differentiate PSP from Parkinsonism. (Oba and others, 2005) as the midbrain is smaller relative to the pons in PSP than in PD.

Recently several simple clinical tests have been reported as being commonly abnormal in PSP, but rarely normal in Parkinsons disease(Kuniyoshi et al, 2002). These include a lack of habituation of blinking after a penlight is flashed 7 times in one eye, a sustained head deviation after unidirectional rotation in an office chair for 45 seconds, and the "applause" sign -- the patient is asked to clap exactly three times, sees it demonstrated, but when they perform the test the clapping continues more than 3 times. Of these, the "applause sign" seems most potentially useful, with the head deviation sign also being a reasonable possibility.

Litvan and others (1997) investigated features that differentiate PSP from these other disorders, as gone over in a table adapted and amended form their paper below. While not to be relied upon as absolute criteria, this analysis does point out that gait instability and supranuclear gaze abnormalities are key  differentiating features.

A syndrome superficially resembling PSP was reported recently in stiff-person syndrome (Oskarsson et al, 2008). Like PSP, this single patient had a vertical supranuclear gaze palsy. Unlike PSP, their saccadic performance fatigued with continuous testing (for 340+ seconds). This patient also developed a tremor. This is an interesting observation. As it is not possible to test most patients for 340 seconds, and in fact most patients with PSP have no oculomotor testing at all, it may be worth checking for GAD antibodies in a prospective fashion in patients carrying the diagnosis of PSP.

Disorder Clinical features of of PSP or disease that are different from other degenerative neurological disorders that have impaired vertical eye movements.
Parkinsonism In PSP there is gait instability, absence of tremor dominant disease, absence of response to levodopa. In PSP there is a supranuclear gaze abnormality
MSA In PSP there is supranuclear gaze palsy, and increased age at symptom onset. MSA prominantly involves blood pressure control.
Corticobasal degeneration Gait instability, severe (supranuclear) upward gaze palsy, bilateral bradykinesia. PET scanning has been reported to differentiate from PSP (Nagahama et al, 1997).
Gaucher's disease

Gauchers has more dementia than PSP, usually starts earlier, and the eye movement problems begin horizontally. Gauchers is rarer than PSP and runs in families.

Picks disease (also known as frontotemporal dementia) There are also many variants of frontotemporal dementia. PSP has more gait instability and less dementia.
Diffuse Lewy body disease (DLB)

PSP has gait instability and supranuclear gaze palsy. Lewy bodies are not increased in PSP (Tsuboi et al, 2001). DLB has a characteristic intolerance of medications.

What is the prognosis of PSP ?

Patients with PSP all progress and the usual life span after diagnosis is 5 years (Bower et al, 1997). The median time from disease onset to first key motor impairment is 4 years, usually 2 years after initial consultation (Goetz, Leurgans et al. 2003). Death does not result as a direct effect of the disease but rather from complications such as pneumonia or pulmonary embolism, which may result from inability of the patients to move about and care for themselves. Of course, falling is common in persons with PSP that are still ambulatory.

Unintelligible speech occurs at a median disease duration of 57 months, or about 5 years. (Goetz, Leurgans et al. 2003)

When persons with PSP begin to cough after every meal, this generally indicates that there is considerable danger of pneumonia from aspiration, and a decision needs to be made whether or not to put in a feeding tube.

Older age at onset and classification as probable PSP are factors associated with poorer survival. Early problems with falls, speech, swallowing, diplopia and early insertion of a gastrostomy tube predicts reduced survival (Nath, Ben-Shlomo et al. 2003)

Is there any proven effective treatment for the cause of PSP ?

In a word, no. In particular, there is no drug known that will stop or reverse the usual inexorable progression of the disease. Most neurologists will try using drugs for Parkinsonism, such as Sinemet. However, such drugs are mildly helpful in only about 50% of persons with PSP. As disability in PSP is due to neuronal damage, and neurons do not regenerate, stopping progression is presently the main goal of treatment.

Logically, it would seem to us and others (Schneider and Mandelkow, 2008) that PSP is caused by tau, and that symptomatic medication is bound to fail. There simply is no reasonable method of regrowing neurons. Thus, prevention of progression should be the goal, and in particular, attention should be directed towards treatments that reduce tau.

Recent studies of specificsymptomatic medications (only the ones in bold seem reasonable to us) include:

There are numerous supportive treatments:

Are there any unproven treatments worth considering ?

There is a study going on at the University of Chicago, of davunetide. This obscure compound is "an active fragment of activity-dependent neuroprotective protein (ADNP)" that "provides protection against kainic acid excitotoxicity in hippocampal neuronal cultures...  It prevents microtubule degradation under oxidative stress and may inhibit the early events of apoptosis" (see  The "" website lists the University of Chicago as the study site in Illinois. 

A recent report (Karapetyan et al, 2013) suggests that astemizole (an antihistamine, Hismanil, withdrawn from the market due to fear of cardiac effects) stimulates autophagy -- basically cell garbage disposal -- in mice. If this also works in humans, astemizole might be useful for a variety of disorders where cellular junk accumulates -- PSP, MSA, and Alzheimer's. Astemizole also is reported to have some anti-tumor activity.

Some neurologists (including the author) prescribe a medication in PSP patients called "seligiline", which was at one point thought to prevent progression in Parkinsonism. Seligiline has not been formally studied in PSP and it is presently unclear whether or not it is helpful. There is some evidence that this drug may inhibit apoptosis (programmed cell death), which might be helpful in PSP. When combined with L-dopa in patients with parkinsonism, this drug is associated with increased mortality (Katzenschlager et al, 2008). It is our view that this drug is worth trying, although not in combination with L-dopa.  It is the authors impression, based on use of this drug in several patients, that seligiline reduces the rate of progression in PSP. Some PSP patients on low dose seligiline develop blood pressure instabilities, which limits its use.

Ropinirole and pramipexole have both demonstrated a reduction in the rate of loss of nigrostriatal innervation as determined by imaging in PD patients, when compared with levodopa. Yamamoto, M. and A. H. Schapira (2008). While this is not the same as showing a neuroprotective effect (as the control was levodopa), nevertheless these agents are worth considering.

Physical therapy has been reported by several groups to have a positive effect (Suteerawattananon et al, 2002; Zampieri and Di Fabio 2008). In general, while we favor physical therapy, it is unreasonable to expect that it will reverse the ongoing degenerative neurological processes that cause the clinical picture of PSP.

There is a small literature that documents a positive effect of amitriptyline (Elavil), which is an antidepressant medication. (Engel 1996). This medication is, however, sedating, drys the eyes, and may also have very adverse effects on thinking.

Drugs like the calcium channel blockers may be also worth trying as they may prevent apoptosis (programmed cell death) as mentioned above for Seligiline. In Mice, nimodipine has been reported as "efficient" in a model for ALS, a related neurodegenerative disorder (Kriz et al, 2003).

Minocycline, a tetracycline type antibiotic, has recently been shown to be neuroprotective in animal models of stroke, multiple sclerosis, Parkinson's disease and Huntington's disease (Arvin et al, 2002). A "cocktail" of minocycline, riluzole (a glutamate antagonist used in ALS) and nimodipine (a calcium channel blocker) was recently reported as "efficient" in mice in a model of ALS, another degenerative neurological disorder (Kris et al, 2003). Memantine (Namendia) is a newly released glutamate antagonist that may also be worth considering. We know of no studies of these drugs in PSP, however.

Flupirtine, an analgesic agent not available in the USA, has several neuroprotective actions including anti-oxident, anti-glutamate, and anti-apoptotic actions. It has recently been found helpful in Creutzfeldt-Jakob disease, which is a degenerative neurological disease similar to "mad cow" disease (Otto M, et al). To our knowlege, it has not been tried in PSP.

Vitamin E may be slightly helpful. In other degenerative neurological disorders, Vitamin E often has a small effect in reducing the rate of progression. It seems possible that agents that reduce oxidative stress (mainly vitamins) might slow the rate of progression in PSP.

The "statin" drugs, used to lower cholesteral also protect against Alzheimer's disease and provide neuroprotection (Chen et al, 2003). These drugs might reduce the speed of progression in PSP. We know of no study of this family of drugs in PSP.

What to do: 


Other Information:

Copyright March 24, 2014 , Timothy C. Hain, M.D. All rights reserved. Last saved on March 24, 2014