The three-step test (also called the Parks-Bielschowsky three-step test or Parks-Helveston three-step test) is a diagnostic method used to identify which cyclovertical muscle is paretic in acquired hypertropia. It was first described by Bielschowsky in 1935 and later systematized and popularized by Marshall M. Parks.
Using the cover test, the deviation is measured in primary gaze, lateral gaze, and head tilt, and the candidate muscle is progressively narrowed down in three steps. In clinical practice, it is most useful for diagnosing superior oblique (trochlear nerve) palsy, but can also be applied to rarer palsies of the inferior oblique or vertical rectus muscles. It also helps differentiate dissociated vertical deviation (DVD) from other vertical strabismus.
Superior oblique palsy is the most common cause of vertical strabismus. Among vertical deviations, fourth cranial nerve (trochlear nerve) palsy is the most common type, with an annual incidence of 6.3 per 100,000 people reported1). The trochlear nerve has the longest intracranial course of all cranial nerves and is the only cranial nerve to exit from the dorsal brainstem, making it susceptible to acquired damage.
This test is designed for the diagnosis of a single vertical muscle palsy. It is less reliable when there are multiple paretic muscles or restrictive strabismus.
Reliability decreases when multiple vertical muscles are simultaneously paretic, in restrictive strabismus, dissociated vertical deviation, or a history of previous vertical muscle surgery. Myasthenia gravis and skew deviation can also cause false positives.
Cycloverical muscle palsy (especially superior oblique palsy) presents with the following symptoms.
Superior oblique palsy (the most common cause of cyclovertical muscle palsy) is classified into the following three types. Reported risk factors include head trauma and aging (decompensation of congenital cases)1).
Congenital
Superior oblique muscle abnormality: Caused by hypoplasia of the superior oblique tendon or abnormal insertion. In more than 70% of cases, the trochlear nerve is absent.
Main complaint is torticollis: From childhood, the head tilts toward the healthy side. To maintain binocular vision with an abnormal head posture, the frequency of amblyopia is low.
No spontaneous recovery: After diagnosis, surgical indication is considered.
Decompensated
Congenital exacerbation: Congenital mild palsy becomes apparent when fusion can no longer be maintained due to aging.
Age of onset: Often occurs in the 20s to 30s with awareness of vertical diplopia. Awareness of cyclodiplopia is rare.
Clues for differential diagnosis: Torticollis can be confirmed in childhood photographs. A wide fusion range (10–15 prism diopters) is characteristic.
Acquired
Traumatic: The most common cause. Bilateral involvement may occur from impact to the midline or top of the head, such as in motorcycle accidents.
Ischemic: Associated with hypertension, diabetes, and hyperlipidemia. Many cases improve spontaneously within 2 to 6 months.
Other: Includes collagen vascular disease vasculitis, hydrocephalus, encephalitis, brain tumor, herpes zoster, and complications after intracranial surgery.
MRI/CT is used to compare abnormalities of the superior oblique muscle insertion and the degree of muscle hypoplasia. In congenital cases, muscle abnormalities are severe, and the traction test shows laxity of the superior oblique muscle. If torticollis is confirmed in childhood photographs, it suggests congenital (decompensated) palsy. In acquired cases, the superior oblique insertion is essentially normal.
Using prisms, the ocular deviation is quantified in primary position, lateral gaze, and head tilt. Small deviations can be measured using prisms with a Maddox rod or red filter.
An overview of each step of the 3-step test is shown below.
| Step | Assessment | Suspected muscles |
|---|---|---|
| Step 1 | Which eye is hypertropic | 2 depressors of hypertropic eye + 2 elevators of hypotropic eye (4 muscles total) |
| Step 2 | Whether it increases with right gaze or left gaze | Narrow down from 4 muscles to 2 muscles |
| Step 3 | Whether it increases with right or left head tilt | Identify 1 muscle from 2 muscles |
In primary position (straight ahead), determine which eye is hypertropic. The depressor muscles of the hypertropic eye (inferior rectus, superior oblique) or the elevator muscles of the hypotropic eye (superior rectus, inferior oblique) are candidates for the paretic muscle. This step narrows down the 8 cyclovertical muscles to 4.
Determine whether the hypertropia increases in right gaze or left gaze. Based on the direction of action of each cyclovertical muscle (the direction where the primary action is maximal), among the 4 muscles remaining from Step 1, the 2 muscles that have the corresponding action in lateral gaze are identified.
Determine whether the hypertropia increases with head tilt to the right or to the left. When the head is tilted, the intorters and extorters are activated, so the paretic muscle can be finally identified among the remaining 2 muscles.
The muscle identified in all three steps is the paretic muscle. If the muscle cannot be identified, consider the possibility of skew deviation.
In the following conditions, the results of the three-step test may be false positive or have reduced reliability.
This test quantifies the cyclotorsional component in addition to the three-step test. It uses two Maddox rod lenses of different colors (red and white).
When the Maddox rods are placed vertically, horizontal lines are seen. In an eye with excyclotorsion, the lines are perceived as oblique. The lenses are rotated until the two lines become parallel, measuring the magnitude and direction of cyclodeviation.
Excyclotorsion exceeding 10 degrees suggests bilateral trochlear nerve palsy. This is a subjective test, with results varying depending on the examiner.
This is an additional test to differentiate skew deviation (vertical misalignment of supranuclear origin) from vertical strabismus due to other causes. Since skew deviation results from an imbalance in otolithic input to the ocular motor nuclei, the deviation decreases when the gravity vector changes in the supine position.
A positive result is defined as a reduction of vertical deviation by 50% or more from upright to supine. Reported sensitivity for skew deviation is 80%, and specificity is 100%1). However, in acute skew deviation (within 2 months of onset), this reduction is not consistently observed, and reliability is reduced1). In trochlear nerve palsy or restrictive strabismus, no significant difference is found between upright and supine positions.
Considering the possibility of skew deviation, add the upright-supine test. Also, MRI to confirm superior oblique muscle atrophy and the presence of the trochlear nerve, and detailed eye movement evaluation using the Hess red-green test or large synoptophore are useful. Exclusion of myasthenia gravis and thyroid eye disease is also necessary.
Treatment for cyclovertical muscle palsy (mainly superior oblique palsy) identified by the three-step test is described. Diplopia, compensatory head posture (sometimes with neck pain), and asthenopia are the basis for treatment 1). Treatment goals are good visual acuity, good binocular vision, and improvement of abnormal head posture.
The selection criteria for surgical procedures are shown below.
| Condition | Recommended Procedure |
|---|---|
| Small angle of deviation | Inferior oblique muscle weakening (partial resection or recession) |
| Hypertropia >15 degrees in primary position | Inferior oblique weakening + additional muscle surgery |
| Torsional diplopia only | Anterior transposition of superior oblique (Harada-Ito procedure) |
| Acquired, mainly torsional diplopia | Recession + nasal transposition of contralateral inferior rectus |
Additional surgical options include superior oblique tuck, superior rectus recession, recession of the inferior rectus in the healthy eye, and anterior transposition of the inferior oblique. The correction amount for excyclotorsion by nasal transposition of the inferior rectus is approximately 6 to 7 degrees per muscle belly.
Because spontaneous recovery is possible, conservative treatment such as prism glasses is generally used for 6 months. Surgery is indicated if diplopia in primary position persists after 6 months. See the “Standard Treatment” section for details.
Step 3 of the three-step test (Bielschowsky head tilt test) is based on the mechanism of compensatory cyclorotation in the vestibulo-ocular reflex.
When the head is tilted to one side, the otolith system sends impulses to the extraocular muscles that compensate for the rotation. For example, tilting the head to the right causes intorsion of the right eye via the right superior oblique and right superior rectus, and extorsion of the left eye via the left inferior oblique and left inferior rectus.
Under normal conditions, the depressing action of the superior oblique muscle and the elevating action of the superior rectus muscle cancel each other out, so no vertical ocular deviation occurs. However, when the superior oblique muscle is paralyzed, the force opposing the elevation of the superior rectus is lost, causing the affected eye to elevate when the head is tilted toward the affected side, increasing hypertropia.
When the head is tilted toward the unaffected side, the superior oblique muscle of the affected eye is not stimulated, so the deviation decreases or disappears. This principle is the basis of the Bielschowsky head tilt test.
The superior oblique muscle originates deep in the orbit, changes direction at the trochlea in the anterior medial orbit, and inserts into the sclera on the temporal side of the superior rectus muscle. The superior oblique tendon attaches broadly like a fan, with posterior fibers primarily responsible for depression and anterior fibers primarily responsible for intorsion. This functional division provides the theoretical basis for surgical procedures such as the Harada-Ito procedure.
The trochlear nucleus is located in the dorsal midbrain, and its nerve fibers run dorsally, decussate in the anterior medullary velum, then pass through the cavernous sinus and superior orbital fissure to enter the orbit. The decussation in the anterior medullary velum is vulnerable to trauma, which is a common mechanism for traumatic bilateral trochlear nerve palsy.
Skew deviation is a vertical misalignment of the eyes caused by supranuclear lesions in the brainstem or cerebellum. It results from an imbalance in otolith input to the ocular motor nuclei. Although it may show a pattern similar to superior oblique palsy on the three-step test, it differs from trochlear nerve palsy in that torticollis occurs toward the hypertropic side and the eye exhibits intorsion. The deviation decreases in the supine position due to a change in the gravity vector, which is the principle of the upright-supine test.
The three-step test is the gold standard for diagnosing cyclovertical muscle palsy, but recent studies have questioned its sensitivity.
A study of 50 patients with MRI-confirmed superior oblique atrophy found that the three-step test failed to detect 30% of superior oblique palsy cases. It was often reported that only two of the three steps were positive.
Another study analyzed sensitivity based on MRI-confirmed presence or absence of the trochlear nerve. Among 166 cases, the diagnostic sensitivity of the three-step test for unilateral superior oblique palsy was 75%.
It has been noted that a positive three-step test does not necessarily indicate cyclovertical muscle palsy. Contracture of the vertical rectus, palsy of multiple vertical muscles, dissociated vertical deviation, previous vertical muscle surgery, skew deviation, myasthenia gravis, and small non-paralytic vertical deviations associated with horizontal strabismus can cause false positives. It has also been reported that abnormalities of the orbital pulley can contribute to a positive Parks three-step test.
To address the issue of inter-examiner variability in bedside subjective tests (red glass test, double Maddox rod test), a computerized red glass test is being developed to quantify and localize diplopia.
The three-step test remains the standard clinical diagnostic method, but given the wide range of differential diagnoses for vertical strabismus, comprehensive evaluation combined with imaging such as MRI is required.
