6.2 The auditory brainstem response (ABR)

The ABR is the averaged scalp potential evoked by a brief acoustic stimulus, recorded over the first 10 ms after stimulus onset, dominated by five consistent peaks named in Roman numerals I through V. Each peak corresponds — with surprising one-to-one fidelity — to a specific stage of the brainstem auditory pathway:

The peaks are vertex-positive (recording electrode at Cz, reference at the ipsilateral mastoid, ground at the forehead). Wave V is the most robust peak in adults and across patients; clinical interpretation centres on its latency, amplitude, and presence at low stimulus levels.

What the ABR is used for

The ABR has two distinct clinical roles that share the same recording but ask different questions of it.

Threshold ABR

The audiologist starts at a high stimulus level (80–90 dB nHL) where wave V is clear, then steps down in 10–20 dB decrements, recording 1500 trials at each level, until wave V can no longer be reliably identified above the noise floor. The ABR threshold is the lowest level at which wave V is present. In normal-hearing adults this is typically 0–20 dB nHL; the ABR threshold correlates with the behavioural threshold for the stimulus’s effective frequency to within 10–15 dB.

The “effective frequency” caveat matters: a click stimulus has a flat broadband spectrum but the cochlea’s response is dominated by the high-frequency basal places (where traveling-wave speed is fastest and synchronisation is tightest). Click-ABR thresholds therefore reflect hearing at roughly 2–4 kHz, not low-frequency hearing. To get frequency-specific thresholds the click is replaced with a tone-burst — a brief (5 ms Blackman-windowed) sinusoidal pulse at a specific frequency. Tone-burst ABR works at 500, 1000, 2000, 4000 Hz; the synchronisation is weaker than for clicks (especially at 500 Hz) and the responses are smaller and need more averaging, so a full tone-burst ABR battery takes considerably longer than a click ABR.

Threshold ABR is the gold-standard objective test in three populations:

• Infants and toddlers who cannot reliably perform behavioural audiometry until age 5–6 months for visual-reinforcement audiometry, and even later for conditioned play. ABR works on a sleeping or sedated child of any age.
• Adults who cannot or will not cooperate with behavioural testing — patients with cognitive disability, malingerers, or those seeking compensation.
• Confirmation of newborn-screen referrals — every infant who refers on the OAE/AABR universal screen receives a diagnostic ABR within 3 months to confirm or rule out the screen finding.

Neurodiagnostic ABR

The same recording, interpreted differently, becomes a neurological localisation test. The key metrics:

• Interpeak intervals. The I–III interval (peripheral transmission) is normally 2.0–2.3 ms; the III–V interval (brainstem transmission) is 1.8–2.0 ms; the total I–V interval is 3.8–4.4 ms. Prolongation of one segment localises pathology to that segment.
• Side-to-side asymmetry. The wave V latency in healthy ears matches between sides to within 0.2 ms; an asymmetry of more than 0.3 ms is a strong indicator of unilateral pathology, most often a retrocochlear lesion.
• Latency-intensity functions. Plot wave V latency against stimulus intensity from 30 to 90 dB nHL. The normal curve is a smooth declining function with a slope of about 0.04 ms/dB (steeper at low intensities). A flat low-intensity portion suggests conductive loss; a uniformly shifted curve suggests cochlear loss; a plateau at long latencies and reduced amplitudes — sometimes with absent waves I–II — suggests retrocochlear pathology.

The retrocochlear ABR pattern

Before MRI became universal, ABR was the screening test for vestibular schwannoma (acoustic neuroma) — a benign tumour of CN VIII that, even at small size, compresses the auditory nerve and slows its conduction. The classical signs:

• Wave V latency prolonged on the affected side by ≥ 0.3 ms relative to the contralateral side.
• I–V interpeak interval prolonged by ≥ 0.3 ms.
• Wave I present but wave V absent or markedly reduced — the lesion is between the cochlea and the brainstem.

Modern MRI has displaced ABR as the primary screen — a small vestibular schwannoma may be MRI-positive but ABR-normal, especially with intracanalicular tumours of < 1 cm. But MRI is expensive and not always accessible; ABR remains useful as a low-cost screen, an intra-operative monitor during posterior-fossa surgery, and a confirmatory test in patients with MRI contraindications.

Recording setup

A clinical ABR system records three channels at minimum:

• Cz – ipsilateral mastoid (A1 or A2): the standard recording channel, optimised for wave V.
• Cz – contralateral mastoid: provides the binaural comparison.
• Forehead ground.

Electrodes are silver-silver-chloride disposable adhesive pads, scrubbed with abrasive prep gel to lower skin impedance below 5 kΩ per electrode. The signal is differentially amplified with gain of about 100,000 (the response is ~0.3 µV; the input voltage to the digitiser needs to be in the tens of millivolts), bandpass filtered 100–3000 Hz, digitised at 25 kHz or higher, and time-locked averaged in software with respect to the stimulus trigger.

Stimulus delivery uses an insert earphone — a tube-coupled transducer inserted into the ear canal with a soft eartip. The insert earphone has two advantages over supra-aural headphones: it provides about 70 dB of interaural attenuation (so cross-hearing is not an issue, eliminating the need for masking in most cases) and it shifts the stimulus arrival by about 0.9 ms (the tube transit time), separating the electrical stimulus artefact from the response — a critical refinement for high-stimulus-level recordings.

Pediatric ABR vs adult ABR

Newborns and infants have ABRs whose absolute latencies are longer (wave V at term: ~7.0 ms at 80 dB nHL, vs ~5.7 ms in adults) and whose interpeak intervals are still maturing. The brainstem myelination process continues for the first 18 months of life; wave V latency reaches adult values around 18 months. Pediatric ABR norms are age-specific: a 3-month-old with a wave V latency of 6.5 ms is normal; the same latency in an 18-month-old is abnormal.

The other practical difference is sedation. An adult can lie still for 60 minutes of ABR recording. A 6-month-old cannot. Pediatric ABR for diagnostic purposes is performed under natural sleep (for infants under 6 months who will sleep cooperatively after a feed) or under sedation (chloral hydrate historically, propofol increasingly). Sedation requires anesthesia involvement and dramatically increases the cost and risk of the test — the major reason ASSR (next lesson) has been pursued so aggressively as a faster, multi-frequency alternative.

The next lesson covers the two younger evoked-potential techniques that complement the ABR: the auditory steady-state response, which provides parallel frequency-specific threshold measurement, and the cortical auditory evoked potential, which probes higher-level auditory processing.