06 Neonatal EEG I: maturation by postmenstrual age

The neonatal brain changes faster than at any later time, and the EEG changes with it, week by week. A pattern that is the expected background at 30 weeks is a serious abnormality at term. So the first act of reading a neonatal record is not to look at the tracing; it is to work out how old the brain is, and the clock that matters is not the calendar.

Postmenstrual age = gestational age at birth + chronological age. A baby born at 28 weeks who is now 6 weeks old is read as a 34-week record. A baby born at term and studied on day 3 is read as a 40-week record. The number goes at the top of the report, before the indication, and every later judgment is made against it, as the reporting order in Module 2 requires.

Why the maturation matters clinically

The reasons this arithmetic comes first are practical. The preterm brain is discontinuous, asynchronous and full of transients that would be alarming in an older child, and a reader who does not know the postmenstrual age will either overcall a normal preterm record or, worse, miss an abnormal term one because it "looks preterm". The features that mature on a schedule, continuity, interhemispheric synchrony, the graphoelements, the sleep states and reactivity, are taught in the following sections, each tied to the weeks at which it appears and disappears. The tables give consensus ranges, and the sources differ by 1 to 4 weeks for most features (André 2010; St. Louis 2016; the Current Practice chapter). Treat them as the shape of normal development, not as cutoffs.

A habit to build now

Before looking at a neonatal tracing, write down the postmenstrual age and say aloud what you expect: continuous or discontinuous, synchronous or not, which graphoelements should be present and which should have gone. Then look. The sign-off for this module reverses the exercise, asking the resident to estimate postmenstrual age from the tracing to within 2 weeks, which is only possible if the expectations were learned in the first place. The same technical points from Module 1 apply: check the page duration, and confirm the state from the ECG and respiration channels, because in a neonate the state changes what the background should look like.

The most striking feature of a very preterm record is that it is not continuous. Activity comes in bursts separated by periods of low voltage, a pattern called tracé discontinu. This is the normal background of extreme prematurity, and it recedes on a schedule. Wake and active sleep become continuous from about 30 to 34 weeks, while discontinuity persists in quiet sleep to about 34 weeks, after which quiet sleep takes on the tracé alternant pattern described later in this module.

How long a gap is allowed

The question a reader has to answer is whether the interburst intervals are too long for the age. The longest normal interburst interval shortens with maturation: about 60 s at 24 weeks, 40 s at 26 weeks, 20 s at 28 weeks, 10 to 20 s at 30 to 34 weeks, about 10 s at 34 to 36 weeks, and about 6 s at term (Learning EEG; consistent with Medscape). Those numbers are not the only ones in print. The French consensus reports wider maxima, for example 12 to 76 s at 24 to 26 weeks (André 2010). Some texts give median rather than maximum intervals, which explains much of the disagreement between sources. When you quote a limit, know which kind of limit it is.

What the numbers are for

The practical use is to separate normal discontinuity from an excessively discontinuous or suppressed background, the abnormal categories defined in Module 7. A 26-week record with 35-second interburst intervals is within the range of normal; a term record with the same intervals is grossly abnormal. Measure the longest intervals, not the typical one, and measure them in quiet sleep, where discontinuity is at its most pronounced at any age. Then compare with the postmenstrual age. A record whose intervals are appropriate for a younger age than the baby's is not "a bit discontinuous"; it is immature or abnormal for age, and the report should say which.

Two technical cautions from Module 1 apply with force here. The neonatal page is often long, which compresses the tracing and makes an interburst interval look shorter than it is: measure against the time marks. And the definition of an interburst interval rests on amplitude, so the sensitivity setting must be known before a period of "low voltage" is called.

Alongside continuity, two other things mature on a schedule: how well the two hemispheres burst together, and which transient waveforms, the graphoelements, are present.

Interhemispheric synchrony

Synchrony follows a curve rather than a line. Before about 27 to 29 weeks the record is hypersynchronous: bursts occur together across the hemispheres. Synchrony then falls, to only about 70% at 29 to 30 weeks, before rising again to nearly 100% by term. The ACNS neonatal terminology defines bursts as synchronous when they occur within 1.5 s of each other (Tsuchida 2013). The clinical use is that asynchrony is expected around 30 weeks and abnormal at term, so the same finding reads differently by age.

The graphoelements

Each of these is a marker of age. Temporal theta is a feature of the very preterm record and should have gone by the early thirties. Delta brushes, fast activity superimposed on slow waves, are the signature of the middle preterm period, most abundant at 31 to 34 weeks, and they should be largely gone by 40 to 44 weeks. Frontal sharp transients arrive later, around 34 to 35 weeks, and are still normal through 44 to 46 weeks; they are blunt, often bilateral, and a resident who has just learned the epileptiform criteria in Module 5 must not score them as spikes. Anterior slow dysrhythmia appears near term, mostly during the shift from active to quiet sleep; no source gave firm age limits for it.

Reading the graphoelements

The presence of a graphoelement past its time is the definition of a dysmature record, treated in the last section of this module. The absence of one at its peak is harder to interpret and should be read with the rest of the background. Ages in the table are consensus ranges, and sources differ by 1 to 4 weeks (André 2010; St. Louis 2016; the Current Practice chapter), so a record one or two weeks off the table is not abnormal; a record six weeks off it is.

By the last weeks before term, the neonatal record organizes into states, and each state has its own background. Reading a term record without identifying the state is like reading an older child's record without knowing whether the eyes were open. The ECG and respiration channels that Module 1 asked for are how the state is identified: active sleep has irregular respiration and eye movements, quiet sleep has regular respiration and little movement, and the wake record is accompanied by the movement and muscle of an awake baby.

Three states, three backgrounds

Tracé alternant is the point at which residents most often confuse normal with abnormal. It is discontinuous, and discontinuity at term sounds alarming. The difference from an abnormal discontinuous background lies in the interburst periods, which in tracé alternant are attenuated but not suppressed and still contain activity, and in the presence of the alternation only in quiet sleep, with a continuous record in wake and active sleep. Module 7 gives the ACNS definitions that draw the line.

Reactivity and the sleep cycle

Two more features mark the maturing record. Reactivity, a change in the background with stimulation, appears from about 30 to 32 weeks and is reliable by 35 to 36 weeks. Its absence at term is an abnormal finding, and one of the things that distinguishes a burst-suppression background from tracé alternant. The sleep cycle also lengthens with age: 30 to 50 minutes below 35 weeks and 50 to 65 minutes above 35 weeks (Tsuchida 2013). The practical consequence is that a recording must be long enough to capture at least one full cycle, including quiet sleep, before the background can be called normal. A 20-minute term record that shows only active sleep has not answered the question.

A dysmature record is one that looks younger than the baby's postmenstrual age: discontinuity that should have resolved, asynchrony that should have gone, graphoelements that should have disappeared, or a sleep organization that has not arrived. It is the single most useful concept in this module, because it turns a table of ages into a finding. The judgment is always relative: the same tracing is normal at one postmenstrual age and dysmature at another, and the only way to make it is to know the age first.

How much deviation counts

The tables are consensus ranges, and the sources differ by 1 to 4 weeks for most features (André 2010; St. Louis 2016; the Current Practice chapter). A record that is a week or two younger than expected is within the uncertainty of the tables themselves, and it is more honest to describe the features and note that they are at the edge of the expected range than to call the record immature. A record that is several weeks younger, with more than one feature out of time, is dysmature, and the report should say so plainly: "the background is immature for the stated postmenstrual age, with features expected at approximately N weeks." The sign-off for this module, estimating postmenstrual age from the tracing to within 2 weeks, is the skill that makes that sentence possible.

What a dysmature record means, and does not mean

Dysmaturity is a description of the EEG, not a diagnosis. It says that maturation of the electrical background lags the stated age. The causes are for the clinician to consider, and they include mistakes in the stated dates as well as illness. The report should state the finding and its degree and leave the cause to the clinical correlation, exactly as Module 2 separates description from interpretation.

Using the free material

None of the three core websites teaches neonatal maturation in depth, so this module leans on dedicated material. The Learning EEG neonatal chapter and its atlas pages show the graphoelements at each age; the Calgary Neonatal Brain Monitoring course walks through the same maturation with tracings; and the AES Neonatal EEG Basics module and the StatPearls chapter give a second pass through the vocabulary. Read one of them with the table from this module open, and check every age you meet against it.