Pull Test + Push and Release Test

Context

Parkinson disease is the second most common neurodegenerative disease in adults after Alzheimer’s dementia (de Lau and Breteler, 2006), affecting around 120,000 people in France (Dorsey et al., 2007) and with an annual incidence of 5–35 new cases per 100,000 people (Poewe et al., 2017). The disease mainly affects people over 60, with a peak of around 65 (Dorsey et al., 2018), and affects twice as many men as women. In economic terms, it represents an estimated annual cost of €13.9 billion in Europe.

Biologically, Parkinson disease is characterized by the progressive disappearance of dopaminergic neurons in the mesencephalic substantia nigra (Redgrave et al., 2010). This neuronal degeneration often begins before the onset of characteristic motor symptoms, underlining the importance of early detection and diagnosis. Indeed, classic motor symptoms often only appear once 50% of the neurons in the substantia nigra have disappeared. It should be noted that some authors include posture among the characteristic symptoms (Dijkstra et al., 2014; Hawkes et al., 2010; Iacono et al., 2015).

During the physical examination and clinical assessment of the Parkinson patient, the aim is to monitor the progression of the disease, identify subjects at high risk of falling according to different situations in daily life, and quantify the effect of a rehabilitation program. It is therefore essential to assess joint mobility, muscle strength, akinesia, tremor, rigidity, and activities, as well as gait and posture disorders.

To assess postural disorders, several clinical tests have been developed to detect the presence and severity of postural instability in Parkinson disease patients. In current practice, the most widely used test for assessing postural instability is the Pull Test or Retropulsion Test. Following the method described by the UPDRS, this test involves the examiner standing behind the patient and pulling back sharply on the shoulders, ready to catch the patient in the event of a fall. In parallel with the Pull Test, some authors have proposed the Push and Release Test, which is based on the Pull Test principle, but takes a slightly different approach.

Video of the Test

Execution

To perform the Pull Test, the patient must stand with eyes open. The examiner stands behind the patient, leaving enough space for an appropriate reaction. The examiner then informs the patient that they must do their utmost not to fall, while remaining ready to catch them if necessary. The examiner then applies a sudden, brief pull on the patient’s shoulders, without giving any specific warning as to when the action is to take place. The test is considered positive if more than three steps are required to regain balance. Beware of freezing events that may occur during the test.

When performing the Pull Test, a number of errors are frequently observed. The patient may be pulled too lightly, compromising balance assessment. The examiner must ensure that sufficient space is left to allow the patient to react and avoid limiting their recovery. Incorrect positioning of the feet, whether too close together or too far apart, can also distort the test. It is essential that the traction exerted is brief and sudden, and not progressive or continuous. In addition, the examiner should pull on the shoulders, not the arms. Finally, the patient should not lean forward before the pull.

To perform the Push and Release Test, the patient stands in a comfortable position, with eyes open. The patient must do whatever is needed to recover balance, including taking one or more steps as necessary.

The examiner stands behind the patient, leaving enough space to allow them to react freely and recover balance independently. The examiner then asks the patient to apply backward pressure against their hands, placed on the patient’s shoulder blades. Suddenly, the examiner withdraws their hands, inevitably forcing the patient to take one or more corrective steps backward to avoid falling.

Clinical Value

Although the Pull Test is part of the routine assessment of patients with Parkinson disease and is widely used in clinical research, its implementation and interpretation are inconsistent. One of the main criticisms of this test is its variability, which can lead to misinterpretation. Frequently observed differences include the state of the patient’s eyes, which may be open or closed, the position of their feet, the presence or absence of precise instructions, and variations in the force or pulling technique exerted by the examiner.

From a statistical point of view, the validity of the Pull Test has been demonstrated, although results vary from study to study. According to Lindholm et al., a score greater than 1 on a Pull Test performed without warning is associated with a risk of falling within six months, with a sensitivity of 0.47 and a specificity of 0.85. According to Munhoz and Teive, a score greater than 2 is correlated with a history of falls, with a sensitivity of 30%, a specificity of 99%, and a positive predictive value of 90%. Jacobs et al. reported a sensitivity of 35% and a specificity of 96% for the detection of falls using this test.

Finally, it is interesting to note that a normal Pull Test could distinguish mild forms of Parkinson disease corresponding to stages 1 and 2 of the Hoehn and Yahr Scale from moderate to severe forms (stages 3 to 5).

The Push and Release Test has also been validated, with results varying from study to study. According to Jacobs et al., the test has a high sensitivity of between 91% and 100%, and a specificity of between 18% and 32%, for detecting patients at risk of falling.

In the study by Valkovic et al., sensitivity was reported to be between 75% and 89%, while specificity ranged from 84% to 97%, with a positive predictive value of between 82% and 96%, depending on the patient’s stage of medication. These results suggest that the Push and Release Test would be particularly effective at identifying Parkinson patients at risk of falling.

In 2016, Jacobs and colleagues also showed that this test displayed a sensitivity of 41% and a specificity of 72% in detecting falls. On this basis, the authors recommend the joint use of the Pull Test and the Push and Release Test to optimize the identification of at-risk patients.

It is also essential to remember that falls result from a complex interaction between several factors, including gait disorders, impaired balance, cognitive decline, and environmental conditions. A Retropulsion Test can, therefore, only assess part of this pathophysiology. This is why the use of balance tests, such as the mini-BESTest or Brief-BESTest, is often recommended for a more comprehensive assessment of fall risk.