INTRODUCTION

Background

An orbital blowout fracture (OBF) is defined as an isolated disruption of the medial and/or inferior orbital wall with an intact orbital rim, resulting from a sudden rise in intraorbital pressure [1]. Diplopia and extraocular movement (EOM) limitation occur in up to one-third of untreated cases and remain key functional endpoints that influence return to work and quality of life [2]. Over the past two decades, advances in high-resolution computed tomography (CT), titanium mesh reconstruction, and evidence-based timing of repair have reduced, but not eliminated, the burden of presentation-time ocular motility deficits [3,4].

OBF epidemiology is strongly shaped by social context. North American emergency department data from 2006–2017 show that violent assault accounted for 43% of orbital floor fractures in young adults, whereas falls dominated in older adults [5]. Taiwanese and Korean nationwide cohort studies similarly report male predominance but divergent age-specific injury profiles that reflect differences in alcohol-related violence and urbanization [6,7].

Functional impairment at presentation is also not explained by fracture size alone. High-energy assaults can force orbital soft tissue and extraocular muscles through the defect, producing immediate diplopia through entrapment. In contrast, low-energy “trapdoor” fractures in children cause transient ischemic paresis that may manifest as early EOM limitation [4,8]. Although a recent meta-analysis demonstrated a threefold increase in persistent diplopia when surgical repair was delayed beyond 14 days [4], the present study focused exclusively on preoperative ocular sequelae, isolating the influence of mechanism, patient factors, and injury era before any operative intervention. To date, no Korean investigation has quantified how a shifting injury spectrum affects these presentation-time sequelae or whether observed center-to-center differences stem from institutional practice or temporal confounding.

Objectives

The objectives of this study were to characterize 13-year temporal changes (2011–2023) in the mechanisms of OBF among patients presenting to two Korean level I trauma centers, to determine how those changes influence the prevalence of preoperative ocular sequelae (diplopia and/or extraocular movement limitation documented before any surgical intervention), and to assess whether intercenter differences in these presentation-time sequelae persist after adjusting for mechanism, demographic, and fracture-specific variables. By integrating a high-volume historical cohort with a recent multiyear sample, this study provides the first Korean evidence that evolving societal patterns of violence are mirrored by shifts in initial functional impairment among OBF patients, independent of surgical management.

METHODS

Ethics statement

This study was approved by the institutional review boards of Inje University Sanggye Paik Hospital (No. 2024-10-011) and Pusan National University Hospital (No. 2406-013-140). The requirement for individual informed consent was waived due to the use of deidentified data and the retrospective design of the study. All study procedures adhered to the principles of the Declaration of Helsinki and followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting guidelines.

Study design and setting

A retrospective, two-center cohort study was conducted. Inje University Sanggye Paik Hospital, a metropolitan level I trauma center serving northern Seoul, contributed all consecutive OBF cases treated in 2011. Pusan National University Hospital, the only level I trauma center in southeastern Korea, contributed all OBF cases treated between January 1, 2019, and December 31, 2023.

Participants

Inclusion criteria were CT-confirmed fracture of the medial and/or inferior orbital wall with an intact orbital rim. Exclusion criteria comprised concomitant facial fractures involving the orbital rim, penetrating ocular trauma, prior orbital surgery, or decreased consciousness that precluded accurate ocular examination. The final analytic sample included 200 patients (161 male and 39 female patients).

Data collection

Demographic and clinical variables retrieved from electronic health records included age, sex, injury date, etiology (categorized as slips, traffic collision, interpersonal violence, low-height falls, bicycle accidents, daily living accidents, or sports), and fracture site (medial, inferior, or extensive). The injury mechanism was classified into seven prespecified categories with explicit operational rules:

(1) Interpersonal violence (assault): deliberate injury inflicted by another person (e.g., punch, kick, or strike with an object) occurring in non-sport or sport settings; if intent was clearly assaultive during sport, the case was classified as assault.

(2) Traffic collision: injuries involving powered road vehicles, either as an occupant or pedestrian (e.g., car, motorcycle, bus, or truck).

(3) Bicycle accidents: pedal-cycle incidents involving the patient as a rider or struck cyclist (e.g., falls from a bicycle or bicycle–vehicle collisions).

(4) Slips: same-level falls attributable to slipping, tripping, or stumbling (e.g., bathroom floor, wet pavement), without a drop from height.

(5) Low-height falls: falls from furniture, beds, chairs, stairs, curbs, or other elevations not exceeding one story.

(6) Sports: injuries sustained during organized or informal athletic activity (e.g., ball, racket, combat, or winter sports) that were not otherwise adjudicated as assault.

(7) Daily living accidents: unintentional impacts not meeting the above criteria (e.g., struck by or against doorframes, tools, or household objects) during routine activities.

(8) Decision rules for ambiguous cases: if both sport and assault were documented, assault took priority; if both slips and sports were documented, sports took priority when the event occurred during play or training—otherwise, it was classified as slips; if both a bicycle accident and traffic collision were documented, the bicycle took priority when the patient was the rider.

The primary outcome was presentation-time ocular motility deficit, defined as diplopia and/or EOM limitation documented at the index presentation prior to any operative intervention.

Statistical analysis

All analyses were performed using IBM SPSS ver. 25.0 (IBM Corp). Continuous variables are summarized as mean±standard deviation or median (interquartile range) and compared using the Student t-test or the Mann-Whitney U-test, as appropriate. Categorical variables are presented as counts (percentages) and compared using the chi-square or Fisher exact tests. Temporal trends across the six study years (2011, 2019–2023) were evaluated using the Cochran-Armitage test for ordered proportions.

Multivariable logistic regression was used to estimate adjusted odds ratios (aORs) for ocular sequelae, incorporating age (per decade), sex, injury mechanism, fracture site, calendar year (continuous), and center as fixed effects. Robust standard errors were clustered by center to account for intrahospital correlation. Linearity of continuous covariates was confirmed using restricted cubic splines. Model performance was assessed with the likelihood-ratio test, Hosmer-Lemeshow goodness-of-fit, and pseudo R². The proportion of missing data was <5% for all variables; therefore, complete-case analysis was applied. Two-sided P-values of <0.05 were considered statistically significant.

Because the historical 2011 series outnumbered the 2019–2023 series, calendar year was modeled as an ordered (continuous) exposure, and center-cluster robust standard errors were used to mitigate disproportionate leverage while preserving representativeness. Cohort-level weighting was considered but not implemented, as each period represented a complete capture for its respective years, and weighting would target a different estimand without improving internal validity. Stratified descriptive rates are presented alongside model-based estimates to maintain transparency.

RESULTS

Patient characteristics and cohort size

A total of 200 patients met the inclusion criteria (161 male and 39 female patients). The historical 2011 cohort comprised 150 cases, and the pooled 2019–2023 cohort comprised 50 cases. Baseline demographics by center are presented in Table 1.

Temporal changes in injury mechanisms

The etiologic profile differed markedly between the two cohorts (χ²=54.5, df=30, P=0.004). Interpersonal violence decreased from 34.7% in 2011 to 14.0% in 2019–2023, corresponding to an 11% annual relative decline on logistic trend analysis (OR, 0.89; 95% confidence interval [CI], 0.81–0.97; P=0.007). In contrast, accidental mechanisms became more prominent: falls increased from 1.3% in 2011 to 12.0% in 2019–2023 and daily living activity–related injuries from 5.9% to 14.0%, with a concurrent rise in traffic collision–related injuries (18.7% to 24.0%), while slip-related injuries showed little change (23.3% to 24.0%) (Fig. 1).

Preoperative ocular motility deficit

Diplopia and/or EOM limitation at presentation was documented in 23 patients (11.5%). The crude prevalence was 14% in 2011 and 4% in 2019–2023 (Z=–1.36, P=0.174). Interpersonal violence (21.8%) and extensive multiwall fractures (21.7%) had the highest unadjusted rates. Male sex (13.0% vs. 5.1%, P=0.048) and age <40 years (15.3% vs. 6.1%, P=0.039) were also associated with higher risk strata (Table 2).

Multivariable predictors of preoperative motility deficit

In the multivariable logistic regression model (pseudo R²=0.129; Hosmer-Lemeshow P=0.470), interpersonal violence emerged as the strongest independent predictor of preoperative motility deficit, increasing the odds of diplopia or EOM limitation by 3.8 times (aOR, 3.84; 95% CI, 1.38–10.65; P=0.010). Male sex demonstrated a similarly strong effect, increasing the odds nearly fivefold (aOR, 4.78; 95% CI, 1.49–15.49; P=0.009). Age showed a protective trend: each additional decade reduced the odds by approximately 25% (aOR, 0.75; 95% CI, 0.56–1.02), although this narrowly missed statistical significance (P=0.064). Compared with extensive multiwall fractures, isolated inferior wall injuries (aOR, 0.61; 95% CI, 0.29–1.80) and medial wall injuries (aOR, 0.49; 95% CI, 0.14–1.32) were associated with lower but nonsignificant risk (both P=0.180). The calendar year variable retained a borderline inverse association (aOR, 0.86; 95% CI, 0.73–1.00; P=0.061), translating to an average 14% annual reduction in preoperative motility deficit after adjustment. Finally, the apparent difference between centers disappeared after full adjustment (aOR, 3.93; 95% CI, 0.93–18.96; P=0.098), indicating that institutional factors themselves did not significantly affect functional impairment at presentation (Table 3). Fig. 2 illustrates these aORs on a logarithmic scale.

Center effect

Unadjusted preoperative sequelae were more frequent at Inje University Sanggye Paik Hospital (14.0%) than at Pusan National University Hospital (4.0%), reflecting the greater violence-related burden in 2011. After adjustment, the center effect lost statistical significance (aOR, 3.93; P=0.098). A sensitivity analysis restricted to male patients aged 20–49 years yielded similar estimates for interpersonal violence (aOR, 4.02; P=0.014) and a reduced center effect (aOR, 1.68; P=0.620), confirming that institutional differences were largely attributable to temporal variation.

DISCUSSION

Key findings

This two-center analysis demonstrates that the decline in interpersonal violence in Korea since 2011 has been accompanied by a marked reduction in preoperative ocular motility deficits after OBFs. Three key observations emerge. First, violence remains the dominant driver of functional impairment at presentation: patients injured by assault had a fourfold higher adjusted risk of diplopia or EOM limitation than those with accidental injuries. Second, the absolute prevalence of preoperative motility deficit fell from 14% in 2011 to 4% in 2019–2023, mirroring an 11% annual contraction in violence-related injuries. Third, the apparent intercenter difference in functional burden disappeared once mechanism, sex, and calendar year were controlled, underscoring that temporal rather than institutional factors account for the disparity.

Comparison with previous studies

Earlier Korean single-center series reported that interpersonal violence accounted for roughly one-third to nearly one-half of OBFs, with preoperative diplopia or EOM limitation present in approximately 18% to 21% of cases [2,9]. Our contemporary cohort showed a violence proportion below 15% and a 4 percentage-point rate of motility deficit, aligning with nationwide administrative data indicating a decade-long decline in assault-related facial trauma [7]. International registries reveal similar trajectories: US emergency department data show a 25% relative decrease in assault-mediated orbital floor fractures between 2006 and 2017 [5], while Taiwanese insurance data document a comparable shift toward falls and low-energy accidents [6]. These convergent findings suggest that societal factors, such as alcohol regulation, public safety campaigns, and demographic aging, are reshaping OBF epidemiology in high-income countries.

Mechanism-specific risk estimates in our preoperative model parallel those from perioperative studies. High-energy assaults typically produce larger herniations and true muscle entrapment, explaining their strong association with diplopia [3]. Age exerted a protective effect, possibly reflecting reduced ocular motility amplitudes and greater orbital compliance in older adults. Male sex increased the odds nearly fourfold, a pattern consistently attributed to greater risk-taking behavior and higher exposure to assault [2,5].

The mid decade “data gap”

National-level data suggest a gradual shift in the mid-2010s in Korea. The 2011–2016 big data from Korea’s Health Insurance Review and Assessment Service (HIRA) showed a steady decline in overall facial fracture incidence with an increasing share of orbital fractures [10]. A multicenter big data study (2015–2020) reported a continuous decrease in surgically treated facial fractures, with falls predominating and personal mobility–related injuries increasing [11]. A national joinpoint analysis of injury mortality and hospitalization (2005–2019) found that assault-related mortality declined continuously (annual percentage change, –6.5%) and assault-related hospitalization decreased more rapidly after 2008 (annual percentage change, –3.76%) [12]. These external findings support a gradual, rather than abrupt, inflection, consistent with the assault-to-accident transition observed between 2011 and 2019–2023 in our cohorts.

Clinical interpretation of the presentation only endpoint

Our endpoint captures functional severity at the initial presentation, which is clinically meaningful because preoperative variables, such as age, fracture displacement, medial wall involvement, preoperative enophthalmos, and timing of repair, are known predictors of postoperative diplopia in contemporary multicenter analyses [13]. Nevertheless, many patients with acute diplopia experience spontaneous improvement without surgery; recent series indicate no new surgical indications beyond day 9 and spontaneous resolution of diplopia in a substantial proportion of follow-up cases [14]. A 2024 systematic review highlighted continued heterogeneity in long-term motility outcomes [15]. Accordingly, the lower presentation-time deficit observed in 2019–2023 should be interpreted as a lower baseline risk, not as a direct estimate of persistent diplopia.

Limitations

Several caveats warrant consideration. First, the retrospective design risks underascertaining mild motility restriction. Second, the 2012–2018 gap prevents within-center characterization of mid-decade dynamics; however, national registries consistently indicate gradual secular declines in facial fracture incidence and assault-coded injuries during this interval [10–12]. Third, the cohort size imbalance (150 patients in 2011 vs. 40 patients in 2019–2023) primarily limits statistical power for temporal contrasts; thus, calendar year was modeled as an ordered exposure, and center-cluster robust standard errors were used rather than post hoc weighting, to preserve the census-like completeness of each period. Fourth, external validity may be region-specific: nationwide data identify rural residence as a risk factor for orbital wall fractures, suggesting that rural or smaller hospitals may exhibit different mechanism distributions and age profiles [7]. Fifth, our outcome was restricted to functional status at presentation; postoperative or long-term motility outcomes were not assessed. These constraints are explicitly acknowledged in our interpretation.

Conclusions

Evolving societal dynamics, particularly the reduction in interpersonal violence, have substantially decreased the functional burden of OBFs at presentation in Korea. Assault mechanism, male sex, and extensive fractures remain potent red flags that should guide early referral and patient counseling. Future prospective registries incorporating operative timing and long-term sensory outcomes will be essential to clarify how modern management further modifies this improving baseline.

ARTICLE INFORMATION

Author contributions

Conceptualization: all authors; Data curation: all authors; Formal analysis: MH, CY; Funding acquisition: CY; Methodology: MH, YK; Project administration: CY; Visualization: MH; Writing–original draft: MH, CY; Writing–review & editing: all authors. All authors read and approved the final manuscript.

Conflicts of interest

The authors have no conflicts of interest to declare.

Funding

This study was supported by a clinical research grant from Pusan National University Hospital in 2022.

Data availability

Data analyzed in this study are not publicly available because they contain sensitive information. Deidentified data are available from the corresponding author upon reasonable request and with appropriate ethical approval.

Fig. 1.

Mechanism distribution.

Fig. 2.

Predictors of preoperative ocular motility deficit.

REFERENCES

• 1. Waterhouse N, Lyne J, Urdang M, Garey L. An investigation into the mechanism of orbital blowout fractures. Br J Plast Surg 1999;52:607–12.ArticlePubMed
• 2. Chi MJ, Ku M, Shin KH, Baek S. An analysis of 733 surgically treated blowout fractures. Ophthalmologica 2010;224:167–75.ArticlePubMedLink
• 3. Boyette JR, Pemberton JD, Bonilla-Velez J. Management of orbital fractures: challenges and solutions. Clin Ophthalmol 2015;9:2127–37.ArticlePubMedPMC
• 4. Damgaard OE, Larsen CG, Felding UA, Toft PB, von Buchwald C. Surgical timing of the orbital "blowout" fracture: a systematic review and meta-analysis. Otolaryngol Head Neck Surg 2016;155:387–90.ArticlePubMed
• 5. Iftikhar M, Canner JK, Hall L, Ahmad M, Srikumaran D, Woreta FA. Characteristics of orbital floor fractures in the United States from 2006 to 2017. Ophthalmology 2021;128:463–70.ArticlePubMed
• 6. Lin YC, Yim CK, Wu AY, Hwang DK. Orbital floor fractures in Taiwan: a 10-year nationwide population-based study. Taiwan J Ophthalmol 2023;13:203–9.ArticlePubMedPMC
• 7. Jung EH, Lee MJ, Cho BJ. The incidence and risk factors of medial and inferior orbital wall fractures in Korea: a nationwide cohort study. J Clin Med 2022;11:2306.ArticlePubMedPMC
• 8. Hsu CR, Lee LC, Chen YH, Chien KH. Early intervention in orbital floor fractures: postoperative ocular motility and diplopia outcomes. J Pers Med 2022;12:671.ArticlePubMedPMC
• 9. Lee YJ. Orbital floor fracture repair with implants: a retrospective study. Arch Craniofac Surg 2021;22:177–82.ArticlePubMedPMCPDF
• 10. Park CH, Chung KJ, Kim TG, Lee JH, Kim IK, Kim YH. Big data statistical analysis of facial fractures in Korea. J Korean Med Sci 2020;35:e57ArticlePubMedPMCLink
• 11. Jun D, Lee JH, Kim SH, Choi JY, Choi J, Lee JH. Big data analysis of facial fracture epidemiology in South Korea from 2015 to 2020: impacts of aging and personal mobility. J Craniofac Surg 2025;36:101–5.ArticlePubMed
• 12. Park HJ, Kim UJ, Lee WK, et al. Joinpoint regression about injury mortality and hospitalization in Korea. J Korean Med Sci 2022;37:e10ArticlePubMedPMCLink
• 13. Hassan B, Hricz N, Er S, et al. Development and validation of a risk calculator for postoperative diplopia following orbital fracture repair in adults. Sci Rep 2024;14:3654.ArticlePubMedPMCPDF
• 14. Soliman L, Sawicki N, Sobti N, Swartz S, Rao V, Woo AS. Re-evaluating the timing of surgery after isolated orbital floor fracture. Plast Reconstr Surg Glob Open 2023;11:e4973ArticlePubMedPMC
• 15. de Santana IH, Viana MR, Palhano-Dias JC, et al. Orbital floor fracture (blow out) and its repercussions on eye movement: a systematic review. Eur J Med Res 2024;29:427.ArticlePubMedPMC

Figure & Data

References

Citations

Citations to this article as recorded by