Introduction

Non-arteritic anterior ischemic optic neuropathy (NAION) is the most common acute optic neuropathy in middle-aged and older adults and a major cause of sudden, painless visual loss and visual field (VF) defects [1,2]. NAION is also the most common cause of acute vision loss related to optic nerve disease. Although the incidence of NAION is relatively high compared with that of other optic neuropathies, the visual prognosis remains unpredictable, and no treatment has been definitively shown to alter its natural course [3–5].

Clinically, NAION is generally associated with an unfavorable visual prognosis, and many patients experience persistent VF defects even after visual acuity stabilizes. Nevertheless, a subset of patients demonstrates partial VF recovery during follow-up. The factors that distinguish these patients from those with persistent visual dysfunction remain incompletely understood. Because no treatment with proven efficacy has been established, understanding the natural history of visual recovery is of paramount importance in NAION [5–7]. From the perspective of both patients and ophthalmologists, knowledge of the expected course of visual function is essential for counseling and management.

In this study, therefore, we aimed to systematically analyze prognostic factors associated with VF recovery in patients with NAION. Recovery of mean deviation (MD) on VF testing was used as the primary outcome for prognostic analysis.

Methods

This retrospective study reviewed the medical records of patients diagnosed with NAION at Pusan National University Yangsan Hospital between January 2017 and December 2024. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Pusan National University Yangsan Hospital (IRB number: 55-2026-062). Due to the study’s retrospective nature, the requirement for informed consent was waived by the Institutional Review Board. The data were accessed for research purposes from March 18 to March 20, 2026.

Patients were included if they had a clinical diagnosis of NAION confirmed by ophthalmologic examination and completed at least 3 months of follow-up. NAION was diagnosed based on established clinical criteria, including acute painless unilateral visual loss, optic disc edema at presentation, optic disc–related visual field defects, and the absence of any systemic or ocular disease that could account for the findings [5,6,8]. Eligibility required comprehensive baseline ophthalmic assessments, including best-corrected visual acuity (BCVA), automated VF testing using a VF analyzer (Humphrey Field Analyzer; Carl Zeiss-Meditec, Dublin, CA, USA), and optical coherence tomography (OCT) performed with the Cirrus HD OCT® (Carl Zeiss-Meditec, Oberkochen, Germany).

Patients were excluded if they had any of the following conditions: (1) arteritic anterior ischemic optic neuropathy, including giant cell arteritis; (2) a history of glaucoma or other optic nerve diseases that could affect VF interpretation; (3) retinal diseases associated with VF defects; (4) neurological disorders involving the visual pathway including cerebral ischemia or infarction confirmed on MRI; or (5) incomplete clinical or imaging data.

Patients who demonstrated an improvement of ≥3 dB in MD at 3 months compared with the initial visit were classified into the recovery group. All statistical analyses were performed using SPSS for Windows (version 28.0; IBM Corp., Armonk, NY, USA). Baseline demographic and clinical variables, including age, sex, visual acuity, and VF parameters, were summarized using descriptive statistics. Comparisons between the recovery and non-recovery groups were performed using independent t-tests and chi-square tests. Correlations among BCVA, VF indices, and retinal nerve fiber layer (RNFL) thickness were assessed using Pearson correlation analysis. RNFL thickness was measured at initial visit using the optic disc cube 200 × 200 scan protocol with the built-in software of the Cirrus HD-OCT (Carl Zeiss Meditec), and average and four-quadrant sectoral values were analyzed. Univariate and multivariate analyses were performed to identify prognostic factors associated with VF recovery. A two-sided P-value < 0.05 was considered statistically significant.

Results

A total of 38 patients (38 eyes) were included in this study. Among them, three patients had bilateral involvement, he first affected eye was selected as the representative eye. Among the 38 patients, 25 were male and 13 were female. The mean age was 63.39 ± 10.82 years (36–84 years). The mean baseline BCVA was 0.55 ± 0.87 logMAR (−0.2 to 3.0), and the mean symptom duration prior to the first visit was 3.29 ± 3.35 weeks (0.5 to 16.0 weeks). Hypertension was present in 20 patients (52.6%), diabetes mellitus (DM) in 7 (18.4%), dyslipidemia in 11 (28.9%), and chronic kidney disease (CKD) in 2 (5.3%). Regarding treatment, 11 patients (28.9%) received Rhopressa (netarsudil 0.02%; Aerie Pharmaceuticals, Bedminster, NJ, USA) and Alphagan P (brimonidine 0.15%; Allergan Inc., USA) eye drops, 14 patients (36.8%) used Alphagan P alone, and 13 patients (34.2%) did not receive any topical neuroprotective eye drops (Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Patient baseline characteristics.

https://doi.org/10.1371/journal.pone.0351219.t001

At baseline, the MD on VF testing was −15.48 ± 9.63 dB (−30.7 to −0.6 dB), the PSD was 9.07 ± 3.55 dB (2.4 to 17.3 dB), and the VFI was 56.84 ± 31.63% (0–99%). The mean RNFL average thickness was 175.63 ± 76.73 μm (70–359 μm), with quadrant measurements of 209.58 ± 93.82 μm (superior; 69–433 μm), 113.71 ± 63.78 μm (temporal; 37–290 μm), 235.50 ± 115.27 μm (inferior; 92–494 μm), and 143.39 ± 78.03 μm (nasal; 53–328 μm).

Ten eyes (26.3%) were classified as the recovery group, and 28 eyes (73.7%) were assigned to the non-recovery group. Baseline BCVA was significantly better in the recovery group than in the non-recovery group (0.21 ± 0.39 vs. 0.68 ± 0.96, P = 0.041). However, no significant differences were observed between the two groups in terms of sex distribution, age, body weight, duration of symptoms, blood pressure, lipid profiles, smoking status, or treatment modality. The location of disc edema also did not differ significantly between the recovery and non-recovery groups (P = 0.227). Similarly, the prevalence of systemic comorbidities, including hypertension, DM, dyslipidemia, and CKD, was comparable between the two groups.

VF parameters indicated less severe impairment in the recovery group, with a higher MD (−11.54 ± 5.93 vs −16.93 ± 10.39 dB, P = 0.059) and higher VFI (72.30 ± 17.58% vs. 51.11 ± 33.94%, P = 0.019). PSD did not differ significantly between the groups. RNFL thickness also showed no significant differences between the recovery and non-recovery groups (Table 2).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Comparison between the recovery and non-recovery groups.

https://doi.org/10.1371/journal.pone.0351219.t002

Pearson correlation analyses were performed to evaluate the relationships between baseline parameters and BCVA at 1, 3, and 6 months. Baseline BCVA demonstrated strong positive correlations with BCVA at 1, 3, and 6 months (r = 0.929, 0.901, and 0.756, respectively; all P < 0.001). Baseline MD and VFI were significantly correlated with BCVA at 3 months (MD: r = −0.458, P = 0.004; VFI: r = −0.545, P < 0.001), whereas PSD showed no significant correlations. Baseline RNFL thickness was not significantly correlated with follow-up BCVA at any time point (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Relationship between baseline BCVA, visual field, RNFL thickness with follow-up BCVA.

https://doi.org/10.1371/journal.pone.0351219.t003

Univariate logistic regression analysis identified no individual variable reached statistical significance for predicting VF recovery (all P > 0.05). However, in the multivariate logistic regression model including baseline BCVA, LDL, MD, and VFI, higher baseline MD (odds ratio [OR], 0.28; 95% confidence interval [CI], 0.09–0.92; P = 0.036), higher baseline VFI (OR, 1.56; 95% CI, 1.05–2.32; P = 0.028), and lower baseline LDL (OR, 0.94; 95% CI, 0.89–1.00; P = 0.050) were independently associated with VF recovery. Baseline BCVA was not independently associated with visual recovery in the multivariate model (Table 4).Other demographic factors, systemic comorbidities, treatment modalities, and RNFL thickness were not independently associated with visual recovery (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Predictive factors for visual field recovery.

https://doi.org/10.1371/journal.pone.0351219.t004

Discussion

The precise pathophysiology of NAION remains incompletely understood, particularly regarding the mechanisms by which ischemia develops at the optic nerve head. Multiple hypotheses have been proposed, including focal arterial insufficiency, systemic hypoperfusion, vasospasm, and dysregulation of optic nerve head microcirculation; however, direct evidence supporting any single mechanism remains limited [2–4,9]. Recent review studies increasingly characterize NAION as a multifactorial disorder rather than the consequence of a single pathogenic process, emphasizing a pathophysiological model involving structural crowding of the optic nerve head and a self-perpetuating cycle of ischemia and edema [4].

Several studies have investigated prognostic factors for visual recovery in NAION. In a large natural history study, Hayreh and Zimmerman reported that patients with milder visual acuity loss and VF defects at diagnosis were more likely to show improvement, with most recovery occurring within the first 6 months after onset [5]. Similarly, the Ischemic Optic Neuropathy Decompression Trial emphasized that the severity of initial VF loss was a key determinant of final visual outcomes, regardless of treatment [10]. Consistent with these findings, patients in the recovery group in the present study had significantly higher baseline MD and VFI values, indicating less severe VF impairment at diagnosis. Moreover, these parameters remained independently associated with VF recovery after adjustment for other variables in multivariate analysis. These results support the concept that the extent of initial functional damage reflects a “functional reserve” that influences the potential for subsequent recovery in NAION [5,10].

NAION is closely associated with systemic vascular risk factors, including hypertension, DM, dyslipidemia, and smoking, which are thought to contribute to microcirculatory dysfunction and impaired autoregulation at the optic nerve head [11–18]. However, the impact of these factors on visual recovery after disease onset has not been consistently demonstrated. In the present study, the prevalence of systemic comorbidities—including hypertension, DM, dyslipidemia, and CKD—did not differ significantly between the recovery and non-recovery groups. These findings suggest that systemic vascular risk factors are important in the pathogenesis of NAION but may have limited value in predicting visual recovery after disease onset. Although LDL levels were lower in the recovery group in univariate analysis, LDL was not an independent prognostic factor in multivariate analysis. Similarly, LDL levels were not significantly associated with long-term visual recovery in a previous study of NAION patients [19]. This may indicate that dyslipidemia influences the microvascular environment; however, the limited sample size and potential confounding factors preclude definitive conclusions.

Age has also been considered a potential prognostic factor in various optic neuropathies. Preechawat et al. reported favorable visual outcomes in NAION patients younger than 50 years [20]. Similarly, Behbehani et al. found that patients younger than 50 years were more likely to achieve better final visual acuity compared with those older than 50 years [18]. In contrast, Sun et al. did not find a significant difference in visual outcomes between patients younger and older than 55 years [21]. In the present study, patient age at diagnosis was not significantly associated with VF recovery in either univariate or multivariate analyses. Further large-scale studies are needed to determine whether age independently influences visual acuity or VF recovery in NAION, or whether, once NAION occurs, the extent of initial functional impairment plays a more critical role in determining recovery potential than chronological age alone.

Previous studies have suggested that the location of disc swelling may influence VF outcomes, as different sectors correspond to distinct retinal nerve fiber bundle distributions and VF regions [22]. In our study, we analyzed whether the predominant location of disc edema affected VF recovery; however, no significant association was identified. This discrepancy from some previous reports may reflect differences in sample size, variability in the timing of fundus evaluation during the acute phase, or the subjective nature of sectoral disc edema classification. Our findings suggest that although the location of disc involvement may influence the pattern of VF defects, it may not independently determine the likelihood of overall VF recovery.

OCT-based analysis of the RNFL is widely used to evaluate structural changes in NAION. In the acute phase, RNFL swelling occurs due to impaired axoplasmic flow, followed by irreversible RNFL thinning secondary to axonal loss [23]. Previous studies have reported inconsistent findings regarding the ability of baseline RNFL thickness to predict functional recovery [21,24]. In the present study, RNFL thickness at diagnosis was not associated with VF recovery. This finding may be explained by the fact that acute-phase RNFL measurements reflect a combination of edema and axonal loss, making it difficult to accurately predict functional prognosis based on structural parameters alone.

Topical eye drop treatment such as Rhopressa(Netarsudil) and Alphagan P(Brimonidine) was not significantly associated with visual field recovery in our study. Brimonidine, an α2-adrenergic agonist, has been suggested to exert neuroprotective effects independent of intraocular pressure (IOP) lowering, potentially through modulation of retinal ganglion cell survival and reduction of excitotoxic damage [25,26]. Similarly, netarsudil, a Rho kinase inhibitor, has been reported to improve optic nerve head perfusion and may influence microvascular dynamics, which are relevant to the ischemic pathophysiology of NAION [27]. Given that NAION is characterized by impaired microcirculation and a compartment-like effect at the optic nerve head, pharmacologic strategies targeting vascular regulation and neuronal survival remain theoretically appealing. However, in the present study, no significant difference in visual outcomes was observed according to treatment groups, which may be attributable to the limited sample size, heterogeneity in treatment initiation timing, and the retrospective nature of the study. Therefore, while our findings do not support a definitive functional benefit of these agents, they do not exclude a potential neuroprotective effect. Prospective, controlled studies with standardized treatment protocols are required to clarify whether such therapies can meaningfully preserve visual function in patients with NAION.

This study has several limitations. First, its retrospective design introduces the potential for selection and information bias. Additionally, variability in the timing of diagnosis and follow-up examinations among patients may have limited the precise characterization of changes in visual function over time. Thus, although patients with known neurological disorders involving the visual pathway were excluded, cerebral ischemic changes, including prior stroke not directly involving the visual pathway or subclinical brain ischemia, were not systematically evaluated. Thus, the potential influence of such factors on visual field recovery could not be assessed in this retrospective study. Second, the relatively small sample size may have limited statistical power, and variables that were significant in univariate analysis may not have been confirmed as independent prognostic factors in multivariate analysis. Further investigations in larger cohorts are needed to clarify the associations between systemic factors, such as HbA1c and lipid profiles, and VF recovery. Additionally, sleep apnea has been recognized as a clinically important risk factor for NAION [28]; however, data on sleep apnea diagnosis were not systematically available in the present cohort due to its retrospective nature. Future prospective studies incorporating systematic sleep apnea screening would help clarify its potential influence on visual recovery in NAION. Third, treatment assignment was not randomized and was based on the clinician’s judgment. Although no treatment has been definitively proven to be effective for NAION, caution is warranted when interpreting the effects of neuroprotective therapies. Additionally, the duration of topical treatment was not uniformly recorded across all patients [29].

Additionally, ganglion cell-inner plexiform layer (GCIPL) thickness, which has been increasingly recognized as an important structural marker in NAION [30], was not included in the analysis due to the limited number of patients with available GCIPL data. Future prospective studies with systematic GCIPL acquisition would help clarify its prognostic value in visual field recovery. Finally, because VF recovery was defined according to changes in MD, functional changes related to central vision may not have been fully captured. Future studies incorporating VF pattern analysis and patient-reported outcome measures may provide a more comprehensive assessment of functional recovery in NAION.

In conclusion, baseline functional status was the most important prognostic factor for VF recovery in patients with NAION. Initial VF parameters, particularly MD and VFI, were independently associated with recovery, whereas demographic factors, systemic comorbidities, and disc edema location were not. These findings suggest that, once NAION has occurred, the extent of initial VF damage may reflect the remaining functional reserve and serve as the most clinically meaningful indicator of recovery potential. Therefore, baseline VF evaluation may provide valuable information for prognostic stratification in clinical practice. Additionally, while topical agents such as brimonidine and netarsudil did not show a significant association with recovery, further research is required regarding the novel agents for the neuroprotection in patients with NAION.