The Romberg sign: A comprehensive clinical re-evaluation for the modern neurologist

Notice: Blog posts are not subject to review by MedLink Neurology’s Editorial Board.

Author: Joaquin A Pena MD

The Romberg sign is one of the most enduring artifacts of the 19th-century neurologic examination, indicating postural instability with eyes closed in a patient who remains stable with eyes open. Initially described in the context of tabes dorsalis, its clinical utility has since expanded to encompass nearly all causes of proprioceptive deficits, from peripheral neuropathies to cervical myelopathies. However, as we move further into the 21st century, the test is the subject of significant debate regarding its specificity, safety, and continued relevance in an era of high-precision diagnostics.

Historical foundations: beyond Moritz Romberg

Although synonymous with the German neurologist Moritz Heinrich Romberg, the sign's history is a tapestry of 19th-century observations. In 1836, the English physician Marshall Hall described a patient who could walk safely while fixating on the ground but stumbled in the dark, noting the patient’s inability to "poise" himself without visual cues. Simultaneously, Bernardus Brach observed that patients with tabes dorsalis were "dependent upon their cane for support" and found walking in the dark nearly impossible.

Romberg’s pivotal contribution in 1840 was to formalize this symptom as a bedside clinical sign. He described the characteristic "oscillations" and subsequent falls that occurred when patients were told to shut their eyes, famously stating that, for these patients, "their eyes are their regulators." It was not until the late 19th century that British neurologist William Gowers definitively linked this phenomenon to dorsal column lesions, thereby transforming the sign from a pathognomonic marker of neurosyphilis into a general indicator of sensory ataxia.

The physiological triad of balance

To interpret the Romberg sign, a clinician must understand that central postural control depends on integrating three peripheral modalities: vision, the vestibular apparatus, and proprioception. Under normal conditions, the functional ranges of these systems overlap, providing redundancy. A healthy individual can maintain stability even if one system is eliminated (eg, by closing the eyes) because the remaining two systems—vestibular and proprioceptive—provide sufficient compensation.

A positive Romberg sign occurs when proprioceptive input is compromised. In these cases, the patient relies heavily on visual feedback (exteroception) to maintain balance with eyes open. When this visual "regulator" is removed by closing the eyes, the remaining vestibular system is insufficient to prevent postural sway or falls.

Proprioception is an "ensemble" of sensory inputs. The muscle spindle is considered the primary receptor, detecting changes in muscle length and velocity. This is supplemented by cutaneous receptors (skin strain) and joint receptors (limit detectors at extreme angles), as well as efferent "outflow" signals from motor commands. Degradation of this complex neurophysiological process can manifest as a positive Romberg sign.

Clinical technique and variations

Despite its ubiquity, there is no standardized approach to administering the Romberg test. Variability exists in:

• Feet position: Whether the feet should be touching (narrow base) or shoulder-width apart.
• Arm position: Held at the sides, extended forward, or extended laterally.
• Duration: Typically conducted for 15 to 30 seconds.

The sharpened Romberg test

To increase sensitivity, clinicians often use the sharpened Romberg test, which requires the patient to stand in a heel-to-toe tandem position. This narrows the base of support, making it more challenging to maintain equilibrium and potentially revealing subtle proprioceptive or vestibular deficits that a standard test might miss.

The walking Romberg sign

Perhaps the most significant modern adaptation is the walking Romberg sign. In this dynamic version, the patient walks 5 meters with eyes open, then repeats the task with eyes closed. A positive Romberg sign is defined as swaying, falling, or an inability to complete the walk due to instability.

Research indicates that this version is significantly more sensitive than the traditional standing test for detecting cervical myelopathy. In a study of 50 patients, the traditional Romberg test was positive in only 34% of cases, whereas the walking Romberg test was positive in 74.5%. Every patient with a positive traditional test also tested positive on the walking version, but 21 patients were identified only by the walking adaptation.

Differential diagnosis: sensory versus cerebellar ataxia

The Romberg sign remains a critical tool for distinguishing between sensory ataxia and cerebellar ataxia.

• Sensory ataxia. Patients are generally stable with eyes open but show marked instability (wobbling or falling) with eyes closed. Their gait often has a "stepping" or "trampling" quality, as they land their feet heavily to maximize sensory feedback.
• Cerebellar ataxia. Patients are often unstable even with their eyes open. Although their sway may increase modestly with eye closure, it does not show the dramatic change seen in sensory ataxia. Their gait is typically broad-based, tottering, and clumsy, resembling acute alcohol intoxication.

Controversy note. Although classic teaching held that the Romberg sign is absent in cerebellar disease, modern posturography shows that patients with vestibulo-cerebellar or spino-cerebellar lesions (specifically in the anterior vermis) exhibit increased sway with eyes closed. Thus, while sensory ataxia yields a more "classic" Romberg, cerebellar ataxia does not strictly rule out a positive result.

Screening for cervical myelopathy

For clinicians screening for cervical spondylotic myelopathy, the combination of the Hoffman sign and the Walking Romberg sign is highly effective. Although the Hoffman sign alone has a sensitivity of only 58% for cord compression, the combination of these two tests—assessing upper- and lower-limb neurologic function—achieves 96% sensitivity for ruling out clinically significant myelopathy. This combination is particularly useful in complex differential diagnoses involving fibromyalgia, carpal tunnel syndrome, or chronic neck pain.

The debate: Is the Romberg sign still relevant?

A modern controversy has emerged regarding whether the Romberg sign should be omitted from routine practice.

The case against: redundancy and risk. Critics, including Martin Turner, argue that the Romberg test "no longer stands up." The primary concerns are:

• Safety. By definition, a positive result is a patient falling to the floor, which poses a risk of physical injury in a litigious era.
• Redundancy. Proprioception can be tested more specifically and safely at the great toe while the patient is seated or lying down.
• Lack of specificity. The test relies on too many integrated networks (vestibular, cerebellar, dorsal columns), making it a poor localizing tool on its own.

The case for: functional insight. Conversely, proponents, including Andrew Lees, argue that the Romberg sign remains essential. Its key strengths include:

• Detecting functional disorders. The Romberg sign can help exclude psychogenic (functional) ataxia; an abrupt, dramatic topple that is absent with eyes open is a valuable semiotic clue.
• Early detection. The Walking Romberg test detects proprioceptive deficits early in the disease process, identifying patients at risk of falling before gross deficits become apparent.
• Global assessment. It provides a "sign-based" overview of axial mobility that isolated limb testing cannot capture.

Innovation: noncontact sensing

Recent advances in neurology have introduced wireless sensing technology (microwave sensing) to quantify the Romberg sign and gait. By using omnidirectional antennas to detect fluctuations in wireless signals caused by body movement, researchers have achieved accuracy exceeding 96% in distinguishing between sensory ataxia and cerebellar ataxia. This technology addresses "self-consciousness" issues, where patients might inadvertently alter their sway when they know they are being watched.

Clinical pearls for practice

• Observe the "bicycle sign." When differentiating Parkinson disease from atypical parkinsonism, ask whether the patient can still ride a bike. Atypical parkinsonism, which involves mediolateral balance deficits, often forces patients to stop cycling early, whereas patients with Parkinson disease usually retain the ability.
• Listen to the gait. Shuffling, flapping, or irregular rhythms can provide diagnostic clues even before the patient enters the examination room.
• Safety first. When performing the Romberg or walking Romberg, always escort the patient or use protective padding to prevent falls.
• Tandem gait as a screen. In early disease stages, when a broad base is not yet apparent, the inability to perform a tandem gait (side-stepping) is a sensitive indicator of instability.

Conclusion

The Romberg sign is not a single, static test but a clinical window into the patient's sensory-motor integration. Although its traditional form may be criticized for limited precision, its dynamic adaptations, such as the walking Romberg test, offer high sensitivity for detecting critical conditions, including cervical myelopathy. The Romberg sign should not be viewed as a redundant relic but as a nuanced component of a comprehensive semiologic evaluation. When combined with adjuncts such as the Hoffman reflex and tandem walking, it provides indispensable diagnostic clarity.

References

Findlay GF, Balain B, Trivedi JM, Jaffray DC. Does walking change the Romberg sign? Eur Spine J 2009;18(10):1528-31. PMID 19387702

Lanska DJ, Goetz CG. Romberg's sign: development, adoption, and adaptation in the 19th century. Neurology 2000;55(8):1201-6. PMID 11071500

Lees A. Replies to “Romberg's sign no longer stands up”: Reply to “Romberg's test no longer stands up”. Pract Neurol 2016;16(5):421. PMID 27630258

Nonnekes J, Goselink RJM, Růžička E, Fasano A, Nutt JG, Bloem BR. Neurological disorders of gait, balance and posture: a sign-based approach. Nat Rev Neurol 2018;14(3):183-9. PMID 29377011

Turner MR. Romberg's test no longer stands up. Pract Neurol 2016;16(4):316. PMID 26951768

Zhang Q, Zhou X, Li Y, Yang X, Abbasi QH. Clinical recognition of sensory ataxia and cerebellar ataxia. Front Hum Neurosci 2021;15:639871. PMID 33867960

Are you interested in contributing a post or becoming a guest blogger for MedLink? Contact us at editorial@medlink.com.