Measurement of the horizon-disc-foveal angle in the assessment of vestibular disorders

Marcello Cherchi, M.D., Ph.D. Page last modified: March 4, 2021

Analysis of eye movements is important in the evaluation of patients with dizziness.  Some of these eye movements are more difficult to study than others.

Everyone knows that the eyeballs rotate.  For instance, just by looking at someone face-to-face, you can detect when the eyeball moves horizontally (left and right) or vertically (up and down), as displayed below.,

The eyeballs can also rotate torsionally, meaning around the visual axis, as displayed below.


Such ocular torsion can occur in normal circumstances.? For instance, when a patient tilts his head left or right, the eyes rotate torsionally in the opposite direction, as if to keep the eyes level with the horizon, as illustrated below.





Sometimes ocular torsion is visible on face-to-face examination.  For instance, in cases of benign paroxysmal positional vertigo (BPPV) an examiner may be able to detect torsional nystagmus by looking closely at the pattern of the iris (the colored part of the eye that usually has a discernible texture).  But if the ocular torsion is static (i.e., not changing), then it cannot be detected on face-to-face examination, so other means are necessary.

Static ocular torsion can be detected by acquiring images of the retina (the back of the eye).  The usual technique for doing this is by identifying two retinal anatomical landmarks (the optic disc and the fovea), drawing a line between them (called the disc-foveal line), and comparing that line to the horizon.  The angle between those two lines (which we call the horizon-disc-foveal angle) has a specific range in normal individuals (see and  The images below are retinal photographs. undus.jpg


In an individual who is upright, it would be abnormal for both eyes to be tilted in the same direction, as displayed below.

Static ocular torsion has been well described in animal models, as well as in humans following surgical lesions such as vestibular neurectomy (see Curthoys and Dai 1991; Curthoys and Halmagyi 1991; Diamond 1981; Halmagyi and Curthoys 1991).

This type of abnormal orientation of the eyeballs, called static ocular torsion, can occur in a variety of disorders, including vestibular neuritis (Vibert et al 1996) and brainstem lesions (Brandt and Dieterich 1987, 1991, 1993, 1994; Choi et al 2014; Dieterich and Brandt 1993a, 1993b; Lee at al 2008; Lee et al 2009; Mossman and Halmagyi 1997).

At Chicago Dizziness and Hearing we are exploring whether the assessment of ocular torsion can help in the diagnosis of vestibular disorders.? In order to accomplish this, we usually refer patients to ophthalmology for a procedure called ocular coherence tomography, which is rapid and comfortable (Brodsky et al 2014).

DFA from Beijing study (Jonas)

Figure is from PLOS-1, and is available online at

Normal values for ocular tilt:

The normal range is ‑7.76° ± 3.63°, with an inter-ocular asymmetry of 4.01° ± 2.94° (Jonas et al, 2011). This is from the Beijing study, where 3052 individuals were studied. The values given are +- 1 SD. A wider limit encompassing about 2 SD would be 0-10.5. As can be seen from the graph above, there are very few individuals with a DFA less than 0, which would mean that the eye is tilted above the horizontal meridian.

Issues with this study is that it was done with Retinal photos (presumably OCT has narrower limits), and also that it was done on a single ethnic group (i.e. Chinese).

Other ways of measuring ocular tilt include subjective vertical, and OCT.