📄 medicine

Estimating height using ulna and tibia measurements in paediatric patients

This study developed and validated height-prediction equations using ulna and tibia measurements, demonstrating high accuracy for healthy children and reasonable performance in ambulatory clinical populations, though systematic overestimation was observed in children with complex conditions.

Original authors: Alessia Salatto, Mary Fewtrell, Susan Hill, Jonathan Wells, Nara Lara-Pompa

Published 2026-06-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Alessia Salatto, Mary Fewtrell, Susan Hill, Jonathan Wells, Nara Lara-Pompa

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine trying to measure the height of a child who is too sick to stand up straight, or whose spine is curved like a question mark. In a hospital, knowing a child's exact height is like knowing the exact size of a key before you try to open a lock; it's crucial for calculating the right amount of medicine or food they need. But if the child can't stand, the usual measuring stick (stadiometer) doesn't work.

This paper is about finding a new, simpler way to guess a child's height using their arm and leg bones as "rulers."

The Problem: The Broken Ruler

Usually, doctors measure height by having a child stand against a wall. But for children with severe illnesses, spinal issues, or those who can't walk, this is impossible. Asking parents to guess the height is like asking someone to guess the weight of a watermelon without a scale—it's often wrong.

The Solution: The "Bone Ruler"

The researchers came up with a clever idea: Your arm and leg bones grow in proportion to your whole body. If you know how long your forearm (ulna) or your shin (tibia) is, you can mathematically guess how tall you are.

Think of it like a tree. If you know how long a specific branch is, you can make a pretty good guess at how tall the whole tree is, even if you can't see the top because it's hidden in the clouds.

How They Did It: Building the Recipe

The team did this in two steps:

Creating the Recipe (The Healthy Group):
They took measurements from 833 healthy children in the UK. They measured the children's actual height, then measured their forearm and shin bones using a simple, non-stretchy tape measure (the kind you might find in a sewing kit).
- The Discovery: They found that if you combine the bone length with the child's age and weight, you can create a "recipe" (a math equation) that predicts height very accurately.
- The Surprise: They thought boys and girls might need different recipes. But once they added "weight" to the mix, the boys and girls fit the same recipe perfectly. It's like realizing that whether you are a small boy or a small girl, the ratio of your arm to your height is the same.
Testing the Recipe (The Hospital Groups):
They took these new recipes and tested them on two groups of sick children at Great Ormond Street Hospital:
- Group A (The "Complex" Group): Children with very serious, complex conditions who were in the hospital.
- Group B (The "Cystic Fibrosis" Group): Children with a lung condition who were generally mobile but nutritionally vulnerable.

What Happened When They Tested It?

1. The Healthy Kids (The Control):
The recipes worked beautifully. The math predicted their height with high accuracy, explaining about 87% of the differences in height just by looking at bone length, age, and weight.

2. The Complex Group (The "Twisted" Tree):
When they used the recipes on the children with complex conditions, the math overestimated their height.

The Metaphor: Imagine trying to measure a tree that has been bent by a storm. The "bone ruler" says, "Based on your branch length, you should be 10 feet tall." But because the tree is bent over, the actual height from the ground is only 8 feet.
The Result: The equations guessed the children were about 4 to 6 cm taller than they actually were when measured. This suggests that for these children, their bones might be normal, but their posture makes them look shorter than they "should" be based on their bones.

3. The Cystic Fibrosis Group (The "Sturdy" Tree):
For this group, the results were much closer.

The Result: The math was very close to the real height, with only a tiny error (less than 1 cm on average).
The Catch: There was a slight pattern: the math guessed the shorter kids were a bit taller than they were, and the taller kids were a bit shorter than they were. It's like a rubber band that stretches a little too much for short people and shrinks a little for tall people.

The "Crowd Wisdom" Experiment

The researchers tried a trick called "Wisdom of Crowds." They took their new recipe, mixed it with an old recipe from another study, and averaged the two guesses.

The Outcome: It didn't really help. It was like asking three people to guess the number of jellybeans in a jar and averaging their answers. Sometimes it got slightly closer, but the range of error (how far off they could be) stayed wide. It didn't fix the fundamental problem of the "bent tree" in the complex group.

The Bottom Line

For Healthy Kids: Using a tape measure on the arm or leg is a fantastic, quick, and accurate way to guess height.
For Sick Kids: It's a useful tool, but you have to be careful.
- If a child has complex posture issues, the math might say they are taller than they really are (because their bones are long, but they can't stand up straight).
- If a child has a condition like Cystic Fibrosis but can stand, the math works quite well.

The Takeaway: This study gives doctors a new, simple tool (a tape measure) to estimate height when the usual method fails. However, just like any map, it's most accurate for the terrain it was drawn for. When the "terrain" (the child's body) is very different due to severe illness, the map might need a little adjustment. The researchers warn that while this is great for groups of patients, using it for a single, very sick child requires caution.

Technical Summary: Estimating Height Using Ulna and Tibia Measurements in Paediatric Patients

Problem Statement
Accurate height (HT) measurement is a cornerstone of paediatric clinical assessment, essential for calculating body mass index (BMI), body surface area (BSA) for drug dosing, and monitoring growth. However, standard stadiometric measurement is often unreliable or impossible in non-ambulatory children, those with skeletal deformities, or patients with neuromuscular disorders. While surrogate methods exist, they often rely on specific subgroups, rigid equipment, or complex imaging (e.g., DXA), limiting their generalisability and practicality in routine clinical settings. There is a need for simple, low-cost, and accurate methods to estimate height in diverse paediatric populations where direct measurement is unfeasible.

Methodology
The study employed a two-phase approach involving the derivation of prediction equations and their subsequent validation in clinical cohorts.

Derivation Phase: Prediction equations were developed using anonymized data from healthy UK children.
- Cohorts: 700 children (aged 4–14 years) provided ulna measurements, and 133 children (aged 10–17 years) provided tibia measurements.
- Measurements: Standing height, weight, and segmental lengths were measured in duplicate using standardised techniques. A non-stretchable tape measure was used for ulna (olecranon to wrist styloid) and tibia (medial knee joint line to medial malleolus) lengths to ensure clinical feasibility.
- Statistical Analysis: Multiple linear regression models were created to predict height. Models tested the inclusion of age, sex, and weight as additional predictors. Goodness of fit was assessed using adjusted $R^2$ .
Validation Phase: The derived equations were tested in two clinical cohorts at Great Ormond Street Hospital (GOSH):
- BodyBasics (BB) Cohort: $n=17$ children with complex conditions and/or chronic diseases, often non-ambulatory or with postural abnormalities.
- Cystic Fibrosis (CF) Cohort: $n=58$ clinically stable, largely ambulatory children with confirmed CF.
- Analysis: Agreement between measured and predicted height was evaluated using Bland–Altman analysis (mean bias and 95% limits of agreement). Correlation coefficients assessed linear relationships, and proportional bias was tested. A "wisdom of crowds" approach was also explored by averaging predictions from the new equations and previously published models (Gauld et al.).

Key Contributions

Development of UK-Specific Equations: The study provides new height-prediction equations specifically tailored for tape-based ulna and tibia measurements in a large, healthy UK paediatric population.
Simplified Measurement Protocol: Unlike previous studies relying on rigid anthropometers or imaging, this work validates the use of a simple, non-stretchable tape measure, enhancing practicality for routine clinical use.
Variable Optimization: The study demonstrates that while sex is a significant predictor in isolation, it becomes non-significant when age and weight are included in the model. This allows for the creation of single, unified equations per bone segment rather than sex-specific ones.
Comparative Validation: The new equations were rigorously compared against measured heights in two distinct clinical populations (complex/chronic vs. ambulatory/nutritionally vulnerable) and against existing models (Gauld et al.).

Results

Derivation Performance: In healthy children, both ulna- and tibia-based equations showed strong predictive accuracy. Including age and weight improved the adjusted $R^2$ $R^{2}$ from approximately 0.79 (bone length alone) to 0.87. Sex was not a significant predictor in the final models.
- Ulna Equation (4–14 years): $HT (cm) = 53.722 + 2.438 \times U + 1.837 \times A + 0.367 \times WT$
- Tibia Equation (10–17 years): $HT (cm) = 58.602 + 2.075 \times T + 1.219 \times A + 0.226 \times WT$
Clinical Validation (BB Cohort): Both equations consistently overestimated height in children with complex conditions (mean bias +4.0 cm for ulna, +6.1 cm for tibia). While correlations were moderate-to-high, the limits of agreement were wide, and no proportional bias was detected, suggesting a systematic overestimation likely due to difficulties in obtaining reliable standing height in this group.
Clinical Validation (CF Cohort): The tibia-based model showed minimal overall systematic error (mean bias +0.8 cm) but exhibited significant proportional bias. The model tended to overestimate height in shorter children and underestimate it in taller children.
Aggregated Predictions: Combining predictions from the new equations with those from Gauld et al. ("wisdom of crowds") reduced mean bias in some instances but did not significantly improve the limits of agreement or eliminate systematic error.

Significance and Claims
The paper claims that tape-based ulna and tibia measurements provide practical and accurate tools for estimating height in healthy children and perform reasonably well in ambulatory clinical populations (such as the CF cohort). The study highlights that these methods are feasible for routine practice due to their low cost and minimal equipment requirements.

However, the authors maintain a modest stance regarding clinical application. They note that the systematic overestimation observed in the BB cohort (children with complex conditions) underscores the need for further validation in larger, more heterogeneous cohorts, particularly those where direct height measurement is impossible. The authors caution that while the equations are useful for group-level assessments, their use for individual patient assessment should be interpreted with care, especially in children with altered body proportions or disease-related growth disturbances. The study concludes that while the new equations generally provide estimates closer to measured height than previous models, they do not fully resolve individual-level imprecision.