What Does Marked Variability Mean in Fetal Heart Tracings?

Marked variability is a pattern seen on fetal heart rate tracings during labor in which the heart rate swings more than 25 beats per minute above and below the baseline. It shows up in roughly 4 to 5 percent of labors and sits in an awkward clinical space: it can reflect a healthy fetus responding vigorously to the stresses of labor, or it can be an early warning that the baby is becoming oxygen-deprived. That ambiguity has made it one of the most debated features in electronic fetal monitoring, with different professional guidelines disagreeing on how long the pattern must last before it counts as abnormal and what, if anything, clinicians should do about it.

What the Tracing Actually Looks Like

During labor, a continuous monitor records the fetal heart rate and prints a tracing that clinicians read in real time. A normal, healthy fetus produces a heart rate that fluctuates gently around a baseline, with swings of about 6 to 25 beats per minute. This moderate variability is generally considered the single most reassuring feature on a tracing because it suggests the baby’s nervous system is intact and responsive. When those fluctuations exceed 25 beats per minute, the pattern is classified as “marked” variability. The tracing takes on a jagged, saw-tooth appearance that can look alarming even to experienced eyes.

The challenge is that different guidelines draw the line differently when it comes to duration. A European research group found that increased variability lasting as little as two minutes can correlate with signs of fetal oxygen deprivation, while the International Federation of Gynecology and Obstetrics (FIGO) guideline from 2015 does not consider the saltatory pattern pathological unless it persists for more than 30 minutes. The UK’s NICE guideline sets its threshold at more than 25 minutes. In one large cohort, episodes lasting longer than 25 minutes were almost nonexistent, appearing in just one case out of nearly 5,000 deliveries.¹ The practical result is that depending on which guideline a hospital follows, the same tracing could be classified as normal, suspicious, or outright pathological.

Why the Fetal Heart Rate Swings So Widely

Before labor begins, the fetal heart rate is regulated by the push and pull of the parasympathetic nervous system (which slows the heart) and the sympathetic nervous system (which speeds it up), shaped by the baby’s sleep-wake cycles and movements.² These two branches of the autonomic nervous system mature at different rates. Before about 32 weeks of gestation, both branches are still ramping up their activity. After that point, the parasympathetic branch becomes dominant during quiet periods, while sympathetic surges drive the heart rate accelerations that clinicians look for as signs of wellbeing.³

During labor, uterine contractions intermittently compress the umbilical cord and placenta, which temporarily reduces blood flow and oxygen delivery to the fetus. When oxygenation drops, chemoreceptors detect the change; when blood pressure shifts, baroreceptors respond. Both trigger reflexes that alter the heart rate.⁴ A fetus that is handling these stresses well may simply produce wider swings as its nervous system responds rapidly to each contraction and recovery. But if the oxygen dips become deeper or more frequent, the same large swings can reflect a nervous system under escalating strain.

Sheep experiments have clarified which branch of the nervous system does the heavy lifting during labor itself. When researchers blocked parasympathetic activity using either the drug atropine or surgical vagotomy, heart rate variability was essentially abolished between contractions, and the heart rate climbed sharply. Sympathetic control, by contrast, took 5 to 10 minutes to recover after each deep heart rate deceleration.⁵ The parasympathetic system, in other words, is the main driver of whatever variability appears on the tracing once labor is well established.

What Marked Variability Means for the Baby

Two recent studies have quantified the relationship between marked variability and newborn outcomes. In a study of more than 8,600 deliveries, about 4.5 percent of tracings showed marked variability before birth. Those babies were not more likely to develop a composite of serious complications overall, but they were roughly 85 percent more likely to have respiratory distress and about 66 percent more likely to have abnormal arterial cord blood gas values, which are a chemical snapshot of how much acid built up during labor.⁶

A second prospective study of nearly 4,400 women looked specifically at marked variability in the 60 minutes before delivery and found it present in about 4 percent of cases. In these births, the risk of neonatal acidosis was roughly doubled.⁷ That is a meaningful signal, but it also means the large majority of babies who displayed marked variability did not develop acidosis. The pattern is better understood as a risk marker that warrants closer attention rather than as a diagnosis of injury.

This distinction matters because fetal monitoring already drives a substantial number of interventions, including emergency cesarean deliveries. Treating every instance of marked variability as an emergency would lead to many unnecessary surgeries. At the same time, dismissing it as benign could cause clinicians to miss the subset of babies who are deteriorating. Navigating that tension is one of the central challenges of modern obstetric monitoring.

The Definition Problem

The clinical utility of marked variability has been hampered by the fact that researchers and guideline bodies have not settled on a single definition. A review of the evidence noted that increased fetal heart rate variability in labor has been “variably defined and poorly understood,” which limits both clinical decision-making and the ability to compare studies.⁸ The term “saltatory pattern” is sometimes used interchangeably with marked variability, while other researchers use “zigzag pattern” to describe a similar-looking tracing with a different minimum duration threshold.

As noted in the FIGO and NICE comparisons above, thresholds for how long the pattern must persist range from one minute to 30 minutes, and the American NICHD system does not specify an exact duration at all. When a Finnish research group studied the zigzag pattern using a two-minute minimum duration, they found it appeared in about 12 percent of deliveries and correlated with higher cord blood markers of oxygen deprivation. But under the FIGO definition requiring 30-plus minutes, the pattern barely existed in their data.⁹ A clinician using one guideline might see a tracing and classify it as marked variability calling for heightened surveillance, while a colleague using a different guideline would call the same tracing normal. This is not a theoretical problem; it plays out on labor wards every day.

Why Clinicians Disagree When Reading Tracings

Even when clinicians use the same guideline, they often disagree on what they see. Research on inter-observer agreement in fetal heart rate interpretation consistently shows that baseline variability is one of the hardest features to classify reliably. A classic study found poor agreement among expert referees when they were asked to classify baseline variability or identify deceleration types on the same tracing.¹⁰ A more recent study using the NICHD three-tier system found that baseline variability had only moderate inter-observer agreement, with a weighted kappa of 0.56, while pathological decelerations were interpreted much more consistently.¹¹

The implication is that marked variability, which depends on accurately judging the amplitude of baseline fluctuations, is particularly vulnerable to subjective interpretation. Two clinicians looking at the same tracing can honestly arrive at different assessments of whether the variability is moderate or marked. This has been one of the driving forces behind the push toward computerized analysis, which at least produces consistent measurements even if debate continues about what those measurements mean clinically.

Computerized Monitoring and Automated Analysis

Since the early 1990s, automated computer analysis of fetal heart rate tracings has offered an alternative to human visual interpretation. The Dawes-Redman system, the most widely used approach, processes the raw signal and extracts features like short-term and long-term variability using standardized algorithms.¹² The computer does not get tired, does not suffer from anchoring bias, and always applies the same criteria. For antepartum monitoring (before labor), computerized systems have become fairly standard in many settings.

During labor, however, the signal is noisier and the clinical context is more dynamic, making automated interpretation harder. Machine learning models trained on expert-labeled tracings have shown accuracy roughly comparable to the median performance of clinical professionals, but not dramatically better.¹³ These tools are promising for standardizing the identification of features like marked variability, but they have not yet eliminated the underlying ambiguity about what the pattern means.

When It Is Not the Baby at All

One hazard that clinicians sometimes overlook is the possibility that the monitor is not actually recording the fetal heart rate. External monitors can inadvertently pick up the mother’s heart rate instead, producing a tracing that may show what looks like marked accelerations and decelerations. A review of perinatal mortality cases reported to the U.S. Food and Drug Administration found unrecognized maternal heart rate artifact in multiple cases where clinicians had been falsely reassured by a seemingly normal tracing.¹⁴

Warning signs include a tracing that suddenly improves from a poor-quality or abnormal pattern to a normal one, accelerations that coincide suspiciously with every contraction (reflecting the mother’s pain or pushing response rather than fetal activity), and baseline rates that seem to fluctuate in character or suddenly jump to a new level. When marked variability or other unusual patterns appear abruptly, confirming that the signal truly comes from the fetus is an essential first step before making clinical decisions based on the tracing.

Lessons from Sheep Experiments

Because researchers cannot ethically subject human fetuses to controlled oxygen deprivation, much of what we know about the physiological meaning of heart rate variability during asphyxia comes from fetal sheep studies. In one set of experiments, preterm fetal sheep exposed to complete umbilical cord occlusion for 20 or 30 minutes showed a distinctive sequence: variability initially increased with the onset of oxygen deprivation, then was suppressed after the occlusion was released. In the 20-minute group, variability gradually recovered over about four hours alongside normal brain activity. In the 30-minute group, variability appeared to recover transiently, but this increase was actually driven by seizure-like electrical activity in the brain and abnormal fetal movements rather than genuine recovery. Once the seizures subsided, variability collapsed and remained suppressed for the remainder of the study period. Histological examination showed severe brainstem injury in the 30-minute group but not in the 20-minute group.¹⁵

A later sheep study reinforced this by showing that lower variability metrics during and after repeated cord occlusions correlated with a greater number of brainstem lesions found on post-mortem examination.¹⁶ The important clinical takeaway from these experiments is that a transient burst of marked variability during or after a stressful event does not automatically mean the nervous system is intact. In severe cases, what looks like recovery on the tracing can actually be the electrical chaos of a brain being injured. Context and trajectory matter far more than any single snapshot.

Variability as a Broader Medical Concept

The idea that fluctuations in a biological signal carry clinical meaning extends well beyond fetal monitoring. Across multiple organ systems, researchers have found that the pattern and degree of variability in a measurement can be as informative as the average level.

Blood pressure is a clear example. Visit-to-visit variability in systolic blood pressure, measured across repeated clinic appointments, predicts cardiovascular events like stroke and heart attack independently of the average blood pressure itself.¹⁷ Someone whose readings bounce between 120 and 170 from one visit to the next may face more risk than someone who sits steadily at 145, even though the average is similar. The swings themselves appear to cause vascular damage.

Blood sugar shows a parallel story in people with type 2 diabetes. Day-to-day swings in glucose levels have been linked to increased oxidative stress, the kind of molecular damage that drives complications like nerve and kidney disease.¹⁸ This has shifted clinical thinking away from focusing solely on long-term average glucose control and toward managing the size of the daily peaks and valleys.

Breathing patterns in intensive care units tell a similar story. In patients on mechanical ventilators, low breath-to-breath variability has been associated with failure to wean off the machine, while more complex, variable breathing patterns tend to predict successful weaning.¹⁹ Models that incorporate heart rate and respiratory variability metrics together have been shown to outperform conventional predictors of weaning success in surgical ICU patients.²⁰

In each of these domains, the principle is the same one that makes fetal heart rate variability so informative during labor: a healthy biological system maintains a dynamic, responsive, somewhat messy signal. Too little variability usually means the system is failing or suppressed. Too much variability, especially in sudden or sustained bursts, can mean the system is being overwhelmed. The clinical challenge is always the same: distinguishing the benign from the dangerous within a signal that is, by nature, never perfectly steady.

How Marked Variability Fits into Clinical Decision-Making

In practice, clinicians do not evaluate marked variability in isolation. The overall context of the tracing matters enormously: Is the baseline rate normal? Are there decelerations, and if so, what type? Has the pattern changed over time or appeared suddenly? Are contractions unusually strong or frequent? What is the clinical situation, such as the mother’s temperature, whether oxytocin is being used, or how far along labor has progressed?

A few minutes of marked variability on an otherwise reassuring tracing in early labor carries a very different meaning than the same pattern appearing after a prolonged run of deep decelerations. The former might simply reflect a vigorous fetal response to a contraction or a period of active fetal movement. The latter is more likely to indicate escalating hypoxic stress and may prompt the team to prepare for operative delivery.

The sheep data reinforces that trajectory is more useful than any single observation. What happens after an episode of marked variability tells the clinician more than the episode itself. If variability settles back to a moderate, reassuring pattern with a stable baseline, the situation is likely benign. If variability drops into a flat, minimal pattern afterward, that suppression is a red flag for potential neurological injury.²¹ And if the tracing shows features suspicious for maternal heart rate artifact, the entire reading may be unreliable and needs confirmation before any decisions are made.

For parents who hear the term during labor, the honest message is that marked variability usually does not mean something is wrong, but it does mean the team is paying close attention. The pattern is common enough that most babies who display it are born healthy, but it raises the level of surveillance because, in a minority of cases, it is one of the earliest signs that the baby needs help sooner rather than later.