Normal findings
What “normal” looks like — and why ±2° is the line most labs draw.
Checking for audio...
The normal range for SVV is narrower than most people realise. Across paradigms and populations, healthy adults set the line within plus or minus two degrees of true vertical.
The upper limit of normal in our reporting is two point zero degrees. Values between two and two point five are borderline. Anything above two point five degrees is abnormal.
Trial-to-trial variability matters as much as the mean. The standard deviation across ten trials in a healthy adult is roughly one degree. A standard deviation above two degrees is abnormal even if the mean is within normal limits.
Age matters slightly. After 70, the normative standard deviation widens to about two degrees. Use age-banded norms if available; otherwise interpret elderly results with that caveat in mind.
Sex effects are negligible. Refractive error and age-related visual change do not bias the SVV provided the bucket or dome is dark enough.
Test-retest reliability is high. The same patient tested twice in one sitting agrees to within about one degree. Reproducibility over weeks is similarly good in healthy subjects but variable in active disease.
Two practical points. First, you can detect cerebellar dysfunction by reporting variability, not mean. Always include the standard deviation.
Second, normal SVV does not mean normal vestibular function. Bilateral vestibulopathy produces a normal SVV because there is no asymmetry to detect. Use the head impulse test and dynamic visual acuity for that.
The normative range
Across a population of healthy adults tested in a dark surround, the mean SVV sits within ±2.0° of true earth-vertical. Test–retest standard deviation within an individual is around 0.7–1.0°[4]. The bucket test agrees with laboratory paradigms to within ~1°[3].
| Population | Mean ± SD | Notes |
|---|---|---|
| Healthy adults, 20–60 y | 0.0 ± 1.2° | Dark surround, laboratory SVV |
| Healthy adults, bucket test | 0.0 ± 1.5° | Mean of 8–10 trials |
| Healthy adults, >70 y | ±0.0 ± 2.0° | Slightly increased variability |
Most laboratories use ±2.0° as the upper limit of normal. Tilts of 2.0–2.5° are borderline and warrant repeat testing; tilts above 2.5° are abnormal[4].
Age effects
Older adults show wider trial-to-trial variability but similar mean tilts. A jump in interquartile range (IQR) in an older patient with a small mean tilt may be more informative than the mean itself — it can signal incipient otolithic dysfunction.
Variability as a localiser
Peripheral vestibular lesions tend to produce a tilted but precise SVV — the patient places the line consistently in the wrong place. Central (especially cerebellar) lesions produce a more variable SVV — the patient’s setting wanders trial to trial. Report the SD or IQR as well as the mean[2].