Abstract
Normative periocular dimensions may vary according to age, gender, and ethnicity. An understanding of normal periocular dimensions is important in the evaluation of orbital disease across different ethnic groups and may serve as reference points for oculoplastic surgery and industry.
Introduction
In clinical practice, "normal" is often a moving target. When a surgeon reconstructs an eyelid after trauma or a clinician evaluates a patient for Thyroid Eye Disease (TED), they require precise standard dimensions to guide their work. These measurements—known as periocular anthropometry—include metrics such as the height of the eyelid opening and the distance between the pupils.
While data exists for specific populations (e.g., Chinese, Turkish, or Korean), Australia’s unique demographic "melting pot" previously lacked its own normative dataset. Furthermore, traditional measurements often rely on manual calipers, which can be inconsistent due to human error and patient movement. This study establishes the first normative database for periocular measurements in a diverse Australian population using digital image analysis, providing a critical reference for oculoplastic surgeons and the ophthalmic industry.
Methods
The researchers conducted a prospective study of 380 participants (760 eyes) at the Royal Adelaide Hospital. To move beyond manual measurement errors, they employed a standardized digital workflow designed for high precision and reproducibility.
Patients were photographed in a controlled environment at a fixed distance. A physical reference—a 24mm green dot—was placed on each participant's forehead to allow software to accurately convert image pixels into millimeters. Using Labelbox, a web-based annotation tool, two reviewers manually marked ten landmarks per eye, including the pupil center, lid margins, and canthi.
The team then integrated the open-source OpenCV library to calculate the final distances between these landmarks. This digital approach ensured high inter-rater reliability, with an Intraclass Correlation Coefficient (ICC) exceeding 0.90 for all measurements, effectively standardizing the baseline for the Australian population.
Results and Key Insights
The study analyzed 760 eyes from a cohort that was 83% Caucasian, 7% East Asian, 7% South Asian, and 2% African, with ages ranging from 18 to 99 years (mean age 58).
The Aging Eye and Sexual Dimorphism
The study confirmed that periocular dimensions change significantly with age, though not uniformly. While the eyelid opening (Palpebral Fissure Height, PFH) and horizontal width (HPA) decrease with age, the distance between the inner corners of the eyes (IICD) actually increases. This suggests that the eye opening narrows and shifts horizontally over time.
Men generally exhibited larger periocular dimensions than women across almost all metrics, including MRD2, PFH, HPA, IICD, IPD, and OICD. However, the upper eyelid position (MRD1) remained relatively similar between genders, indicating that the vertical position of the upper lid relative to the pupil is a stable metric across sexes.
Ethnic Signatures and Clinical Symmetry
Ethnicity creates distinct anatomical profiles even within a single national population. For instance, East Asian participants showed significantly larger IICD compared to Caucasians, while African participants had wider-set pupils (IPD).
Crucially, the study found no significant difference between a person's right and left eye measurements. Consequently, any major unilateral asymmetry in a clinical setting should be viewed as a potential indicator of pathology, such as a tumor or inflammatory disease.
Quantitative Metrics
- MRD1 (Upper lid to pupil): Mean of 3.5 mm. It showed a negative correlation with age ($r = -0.09, p = 0.01$), likely due to age-related weakening of the levator muscle.
- PFH (Palpebral fissure height): Mean of 8.7 mm. This was significantly higher in males (9.0 mm) than females (8.5 mm).
- HPA (Horizontal width): Showed the strongest negative correlation with age ($r = -0.39, p < 0.01$), suggesting the eye opening narrows significantly over time.
Discussion and Limitations
While this study provides a robust baseline, several constraints exist. The small sample size of certain ethnic subgroups, such as African participants ($n=9$), may have masked some statistically significant differences. Additionally, the cross-sectional design compares different age groups at one point in time rather than following individuals longitudinally, which would more accurately capture the aging process.
Future research should focus on correlating these photographically derived measurements with "gold standard" clinical measurements and exploring longitudinal changes. This standardized, digitally-derived baseline represents a significant step forward in improving surgical outcomes and disease diagnosis in the Australian context.
Summary updated