Journal of Pediatric Orthopaedics - 2026-05-20 - Journal Article
Is the Greulich and Pyle Bone Atlas Suitable When Measuring Bone Age for Pediatric ACL Reconstruction in the Modern US Population? Analysis Within the PLUTO Prospective Multicenter Cohort Study.
Guevel B, Hutchinson L, Miller P, Anderson CN, Busch MT, Chambers HG, Christino MA, Cordasco FA, Edmonds EW, Fabricant PD, Ganley TJ, Green DW, Heyworth BE, Lawrence JTR, Matava MJ, Micheli LJ, Milewski MD, Nepple JJ, Parikh SN, Pennock AT, Perkins CA, Saluan PM, Shea KG, Wall EJ, Willimon SC, Kocher MS
Topics
Key Takeaway
Greulich and Pyle bone age overestimates chronological age by a mean of 2.9 months cohort-wide, with ~2-year 95% limits of agreement, but shows no systematic bias across race, ethnicity, or socioeconomic status in 472 pediatric ACL patients.
Summary Depth
Choose how much analysis to show on this article page.
Summary
This retrospective subanalysis of the PLUTO prospective multicenter cohort assessed agreement between chronological age and G&P bone age in pediatric ACLR candidates using Bland-Altman analysis, stratified by race, ethnicity, sex, BMI, and area deprivation index. G&P overestimated chronological age by 2.9 months (SD 12.8) with 95% limits of agreement spanning ~2 years; significant age-strata biases were found at ages 11–13 and 15. No systematic bias was detected across race, ethnicity, or socioeconomic quartiles, but female sex conferred 3.9 months greater bias and underweight BMI produced 15.9 months lower bone age versus chronological age.
Key Limitation
The ~2-year 95% limits of agreement mean that for any individual patient—particularly those near the physeal-sparing decision threshold—G&P alone carries substantial uncertainty regardless of demographic group.
Original Abstract
BACKGROUND
A reliable measure of bone age is of critical importance when planning anterior cruciate ligament reconstruction (ACLR) in a pediatric population. The Greulich and Pyle (G&P) method is still the most common method for bone age determination; however, it is based on radiographs from healthy Caucasian children of high socioeconomic status (SES) in Cleveland, Ohio, in the 1920s to 1940s and may not reflect the modern diverse United States population. The aim of our study was therefore to assess the degree of agreement between chronological age and bone age measured using G&P and further investigate whether there are any systematic differences across race, ethnicity, or SES.
METHODS
This was a retrospective subanalysis of PLUTO (Pediatric
ACL
Understanding Treatment Options), a prospective multicenter cohort study investigating outcomes in pediatric ACLR. All patients with bone age scores using the G&P method preoperatively were included. Data on race, ethnicity, biological sex, BMI, and SES [using a composite marker known as the area deprivation index (ADI)] were collected. Bland-Altman analysis was used to assess the agreement between chronological-measured and G&P-measured age. Systematic bias was defined as a significant mean difference between chronological-measured and G&P-measured age and was determined using Student t test for the cohort and within strata; and differences across strata were determined using a one-way ANOVA. P-values across strata were adjusted using the method of Benjamini and Hochberg to control the false discovery rate.
RESULTS
A total of 472 patients were included in the final analysis. Chronological age at evaluation was lower than the determined bone age for the cohort with a systematic bias of 2.9 months (SD, 12.8) (P<0.001). The Bland-Altman 95% limits of agreement between bone age and chronological age were ∼2 years. There was significant variation in agreement in those aged 11 to 12 (P<0.001), 13 (P=0.008), and 15 years (P=0.008). Female patients had a 3.9-month greater bias compared with males (P=0.003). Underweight patients exhibited nearly 15.9 months lower bone age versus chronological age (P=0.008). No systematic biases were detected across ethnicity (P=0.21), race (0.13), or ADI index quartiles (P=0.23).
CONCLUSION
Our study supports the continued use of G&P for bone age assessment across different racial, ethnic, and sociodemographic groups, with no significant differences detected across these strata. The observed ∼2-year limits of agreement between G&P bone age and chronological age reinforce the importance of skeletal age assessment in this cohort, particularly during the peripubertal period where large differences between chronological and skeletal age can exist. Systematic biases were found in certain age strata, between sexes, and in patients with low BMI. Clinicians should be aware of these variations and consider using multiple methods of assessing skeletal maturity when making growth-sensitive clinical decisions.
LEVEL OF EVIDENCE
Level III-retrospective cohort.