Introduction

Choroidal thickness is increasingly recognized as a critical biomarker in the pathophysiology of retinal disease. Recent advances have highlighted its role not only in the pachychoroid disease spectrum but also in other major retinal conditions, including age-related macular degeneration (AMD), diabetic retinopathy (DR), and high myopia [1–3]. With the development of enhanced depth imaging and swept-source optical coherence tomography, choroidal thickness can now be assessed with high reproducibility in vivo, enabling large-scale epidemiologic investigations [4].

Choroidal thickness is reported to be influenced by demographic and ocular factors such as age, sex, and axial length, while the impact of lifestyle behaviors remains inconsistently reported [5–9]. Among these, cigarette smoking is a major modifiable risk factor for systemic vascular disease and has been consistently associated with a range of retinal diseases such as AMD and central serous chorioretinopathy (CSC) [10,11]. These associations suggest a potential role in modulating disease pathophysiology through alterations in choroidal structure and function. Prior studies on the effects of smoking on choroidal thickness have reported conflicting results, and evidence remains limited due to small sample sizes and insufficient statistical power [12–14]. Notably, there is a paucity of large-scale study has evaluated differences in choroidal thickness across never-smokers, former smokers, and current smokers.

Given the public health burden of smoking and the emerging recognition of choroidal thickness as a clinically relevant biomarker in ocular disease, clarifying their association is necessary. Accordingly, we investigated the association between smoking status and choroidal thickness in a general adult population using data from the Korea National Health and Nutrition Examination Survey (KNHANES) 2017–2021, a nationally representative dataset, thereby providing robust population-level evidence.

Materials and methods

Results

A total of 38,678 participants from the 2017–2021 KNHANES data were initially screened. After exclusion of participants without ophthalmic examinations or with ocular pathologies potentially influencing choroidal thickness, 22,130 individuals were included in the final analysis (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow diagram of participant selection from 2017–2021 KNHANES.

Among 38,678 participants, individuals were excluded due to lack of ophthalmic examination or survey weights (n = 6,326), presence of ocular diseases including diabetic retinopathy, age-related macular degeneration, epiretinal membrane, macular hole, or retinal vein occlusion (n = 4,387), or missing smoking data (n = 5,835). A total of 22,130 participants were included in the final analytic sample, comprising never smokers (n = 13,622), former smokers (n = 4,586), and current smokers (n = 3,922).

https://doi.org/10.1371/journal.pone.0358964.g001

The mean age of participants was 47.0 ± 0.2 years, and 50.1% were women (Table 1). Age and sex distribution differed significantly among the three smoking groups (both p < 0.001). BMI, blood pressure status, DM, cardiovascular disease, and dyslipidemia also showed significant differences across smoking groups (all p < 0.001). Regarding ocular parameters, AL and CMT in both eyes differed significantly among the smoking groups (both p < 0.001). Mean SFCT differed significantly among the groups, showing a progressive increase from never to former to current smokers in both eyes (right eye: 297.3 ± 1.3 µm, 307.0 ± 2.1 µm, and 330.6 ± 2.3 µm, respectively; left eye: 297.2 ± 1.3 µm, 305.5 ± 2.1 µm, and 330.1 ± 2.2 µm, respectively; both p < 0.001). Fig 2 illustrated the weighted mean SFCT according to smoking status, showing a clear stepwise increase from never to former to current smokers in both eyes.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Demographic and clinical characteristics of study participants by smoking status.

https://doi.org/10.1371/journal.pone.0358964.t001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Weighted mean subfoveal choroidal thickness (SFCT) for the right and left eyes is shown across smoking status (never, former, and current smokers).

The solid line represents the right eye and the dashed line represents the left eye. Error bars indicate survey-weighted standard errors.

https://doi.org/10.1371/journal.pone.0358964.g002

In univariate analysis, both former and current smokers demonstrated significantly greater SFCT compared with never smokers (right eye: β = 9.78 ± 2.29 µm and 33.37 ± 2.62 µm, respectively; left eye: β = 8.26 ± 2.32 µm and 32.88 ± 2.47 µm, respectively; all p < 0.001) (Table 2). In multivariate analysis, current smoking remained independently associated with thicker SFCT (right eye: β = 17.55 ± 3.45 µm; left eye: β = 20.47 ± 3.40 µm; both p < 0.001). Other independent factors associated with thinner SFCT in both eyes included older age (right eye: β = −2.13 ± 0.12 µm; left eye: β = −2.04 ± 0.16 µm; both p < 0.001), female sex (right eye: β = −22.60 ± 2.98 µm; left eye: β = −18.91 ± 3.32 µm; both p < 0.001), longer AL (right eye: β = −23.32 ± 2.59 µm; left eye: β = −22.04 ± 3.87 µm; both p < 0.001), greater CMT (right eye: β = −0.17 ± 0.05 µm, p = 0.001; left eye: β = −0.15 ± 0.06 µm, p = 0.006), and the presence of cardiovascular disease (right eye: β = −15.50 ± 6.66 µm, p = 0.020; left eye: β = −18.29 ± 7.58 µm, p = 0.016).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Linear regression analyses of factors associated with subfoveal choroidal thickness according to smoking status.

https://doi.org/10.1371/journal.pone.0358964.t002

Among current smokers, no significant association was observed between cumulative smoking exposure, as quantified by pack-years, and SFCT in the right eye (β = −0.27 ± 0.19 µm per pack-year, p = 0.159), whereas a weak but statistically significant positive association was noted in the left eye (β = 0.41 ± 0.20 µm per pack-year, p = 0.040) (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Linear regression analyses of associations between cumulative smoking exposure and subfoveal choroidal thickness in current smokers.

https://doi.org/10.1371/journal.pone.0358964.t003

Discussion

In this nationally representative Korean population, current smoking was independently associated with greater SFCT in both eyes, whereas former smoking showed no significant association after adjustment for confounders. Nevertheless, a stepwise increase in SFCT was observed across smoking status from never to former to current smokers.

Previous studies investigating the relationship between smoking and choroidal thickness have reported conflicting results. Ulaş et al. reported a transient increase in choroidal thickness within 5 minutes after smoking, which returned to baseline by 60 minutes, suggesting an acute vasodilatory response [13]. In contrast, Sizmaz et al. found a significant decrease in choroidal thickness within 1 and 3 hours after smoking, suggesting opposing short-term vasomotor responses [12]. More recently, Alhazmi et al. showed that nicotine exposure alone, whether via gum or electronic cigarettes, did not significantly alter choroidal thickness, indicating that other components of cigarette smoke may contribute to vascular changes [22]. A meta-analysis including 13 observational studies also found no consistent long-term association, although regional differences were noted [14]. Compared with these studies, which were limited to small or short-term cohorts, the present investigation leveraged a large, nationally representative population with clear classification of smoking status into never, former, and current smokers, enabling robust assessment of chronic exposure.

Cigarette smoke induces both sympathetic stimulation through nicotine and vasodilatory effects through carbon monoxide (CO), resulting in transient alterations in choroidal perfusion and vascular tone [23,24]. Moreover, smoking is known to decrease nitric oxide (NO) bioavailability, which impairs endothelium-dependent vasodilation and may lead to compensatory vascular responses in the choroid [25]. These CO- and NO-mediated changes may explain the thicker SFCT observed in current smokers. With continued exposure, chronic smoking may further provoke compensatory vascular remodeling in response to sustained hypoxic stress, inflammation, and endothelial dysfunction, as demonstrated in systemic vascular studies [26,27]. Even after cessation, partial persistence of these vascular adaptations could account for the marginally thicker SFCT observed among former smokers, suggesting a residual effect of prior exposure.

In addition, the present findings align with evidence linking smoking to pachychoroid-spectrum diseases, particularly CSC. Smoking has been consistently identified as a major risk factor for CSC, with proposed mechanisms including sympathetic activation, oxidative stress, and choroidal vascular hyperpermeability [11,28]. Given that CSC is characterized by choroidal thickening and vascular congestion, the greater SFCT observed in current smokers in our study parallels this pachychoroid phenotype.

Among current smokers, cumulative smoking exposure quantified by pack-years showed only a weak positive relationship with SFCT in the left eye, without a clear dose-dependent trend. A ceiling effect may contribute, as choroidal thickening cannot increase indefinitely with greater smoking duration or intensity, potentially blunting the observable dose–response relationship with pack-years. This pattern also suggests that choroidal thickening associated with smoking likely reflects changes that occur in the early phase of smoking exposure rather than slowly progressive structural remodeling with cumulative lifetime exposure.

Several limitations should be acknowledged. First, the cross-sectional nature of the KNHANES data precludes causal inference and limits the ability to assess longitudinal changes in choroidal thickness over time. Second, although major systemic and ocular confounders were adjusted for, unmeasured or unknown factors influencing SFCT may still exist. Third, smoking exposure was self-reported, potentially subject to recall bias. Fourth, the study population was limited to Korean participants, which may restrict the generalizability of our findings to other ethnic populations. Finally, because the timing of OCT imaging could not be standardized, diurnal variation in choroidal thickness and differences in the interval between the last cigarette and image acquisition may have introduced additional variability.

In conclusion, current smoking was associated with greater SFCT in a large, nationally representative Korean population. These findings suggest an association between smoking status and choroidal structural alterations, although longitudinal studies are needed to determine whether these changes contribute to the development or progression of retinal diseases. Recognition of smoking as a modifiable factor associated with choroidal thickness highlights its clinical relevance and reinforces the broader public health importance of smoking cessation.

Acknowledgments

No additional contributors outside the listed authors met the criteria for acknowledgment.