Cross-sectional associations between variations in ankle shape by statistical shape modeling, injury history, and race: the Johnston County Osteoarthritis Project

Amanda E. Nelson, Yvonne M. Golightly, Shahmeer Lateef, Jordan B. Renner, Joanne M. Jordan, Richard M. Aspden, Howard Hillstrom, Jennifer S. Gregory

Research output: Contribution to journalArticlepeer-review

12 Citations (Scopus)
10 Downloads (Pure)


Injury is an important risk factor for osteoarthritis (OA), a highly prevalent and disabling joint disease. Joint shape is linked to OA, but the interplay of injury and joint shape and their combined role in OA, particularly at the ankle, is not well known. Therefore, we explored cross-sectional associations between ankle shape and injury in a large community-based cohort.

Ankles without radiographic OA were selected from the current data collection of the Johnston County OA Project. Ankles with self-reported prior injury were included as injury cases (n = 108) along with 1:1 randomly selected non-injured ankles. To define ankle shape, a 68 point model on weight-bearing lateral ankle radiographs was entered into a statistical shape model, producing a mean shape and a set of continuous variables (modes) representing variation in that shape. Nineteen modes, explaining 80% of shape variance, were simultaneously included in a logistic regression model with injury status as the dependent variable, adjusted for intra-person correlation, sex, race, body mass index (BMI), baseline OA radiographic grade, and baseline symptoms.

A total of 194 participants (213 ankles) were included; mean age 71 years, BMI 30 kg/m2, 67% white and 71% women. Injured ankles were more often symptomatic and from whites. In a model adjusted only for intra-person correlation, associations were seen between injury status and modes 1, 6, 13, and 19. In a fully adjusted model, race strongly affected the estimate for mode 1 (which was no longer statistically significant).

This study showed variations in ankle shape and history of injury as well as with race. These novel findings may indicate a change in ankle morphology following injury, or that ankle morphology predisposes to injury, and suggest that ankle shape is a potentially important factor in the development of ankle OA.
Original languageEnglish
Article number34
JournalJournal of Foot and Ankle Research
Publication statusPublished - 26 Jul 2017

Bibliographical note

Rheumatology Research Foundation Medical Student Preceptorship Award
(Lateef/Nelson), NIAMS K23 AR061406 (Nelson); NIH/NIAMS P60AR064166
and U01DP003206 (Jordan/Renner), NIH/NIAMS R01AR067743 (Golightly). The
funders had no role in study design; collection, analysis, or interpretation of
data; writing the manuscript or the decision to submit for publication.


  • ankle
  • injury
  • joint shape
  • radiography
  • racial differences


Dive into the research topics of 'Cross-sectional associations between variations in ankle shape by statistical shape modeling, injury history, and race: the Johnston County Osteoarthritis Project'. Together they form a unique fingerprint.

Cite this