Parents are handed a growth chart at routine checks and usually read the percentile as a score. The chart is a description of position within a population, not a measure of health.
A percentile is a rank, not a grade
The curves on a chart are drawn from measurements of many children. A child on the tenth percentile for weight is simply smaller than most children of that age and sex.
Someone has to be on every line for the chart to exist. Being below the middle is not evidence of a problem any more than being above it is evidence of thriving.
Height percentile in particular is strongly influenced by parents' own heights, which is why some services record parental heights and compare a child against that expectation.
The shape of the line matters more than the point
What clinicians watch is whether a child tracks along a curve over time. Growth is a rate, and a single measurement contains no information about rate at all.
A child who has followed one curve steadily and then crosses downward through several is more interesting clinically than a child who has always sat near the bottom of the chart.
This is why repeat measurements at intervals are requested rather than one careful measurement. The value of the chart accumulates with the number of points on it.
Standards and references are different documents
Some charts describe how children in a defined population actually grew, and are references. Others describe how children grow under conditions considered favourable, and are standards.
A child can therefore appear on different percentiles depending on which chart the clinic uses, particularly in infancy, without anything about the child having changed.
Charts specific to certain conditions, and adjustments for babies born preterm, also exist. Comparing readings taken from different charts produces apparent changes that are artefacts.
Measurement error is larger than parents expect
Measuring a wriggling toddler's length is difficult, and small errors translate into visible movement on the chart. Head circumference and length are especially prone to this.
Weight varies with clothing, nappies, time since feeding and time of day. A gain or loss recorded between two visits can reflect the conditions of measurement rather than the child.
Clinicians treat single anomalous points cautiously for this reason, and often remeasure before acting. A parent looking at the same dot has no way to know which kind of point it is.
When a crossing is worth attention
Sustained crossing of curves, a large gap between weight and height percentiles, or a plateau in a period when growth should be rapid are the patterns that prompt further assessment.
Those patterns are interpreted alongside feeding, illness, family history and development, none of which appear on the chart itself. The chart is one input among several.
Interpreting a child's growth is a clinical task, and concerns about the trajectory belong with a paediatrician or health visitor rather than with comparisons against other children of the same age.