Over recent decades, the incidence of allergic conjunctivitis (AC) has steadily increased, with prevalence estimates ranging from 6% to 30% in the general population (Leonardi et al., 2015). AC is characterized by symptoms such as itching, redness, a whitish conjunctiva, and sometimes thick mucous discharge, severe cases may present with papillary hypertrophy of the palpebral conjunctiva (Friedlaender, 1993). Notably, AC is recognized as a key risk factor for inducing dry eye disease (DED) (Villani et al., 2018), significantly impacting patients’ quality of life. The meibomian glands (MGs), located within the eyelids, play a crucial role in maintaining the stability of the tear film by secreting its lipid layer, thereby reducing tear evaporation (Dietrich et al., 2021). Recent studies have shown that AC can contribute to meibomian gland dysfunction (MGD) (Mizoguchi et al., 2017a), leading to tear film instability and exacerbation of ocular surface discomfort.

Previous research has highlighted the potential impact of allergic conjunctivitis on MG morphology. Arita et al. (2012) suggested that allergic reactions might be the reason for increased MG deformation in patients with contact lens-associated allergic conjunctivitis. Similarly, Liu et al. (2022) proposed that morphological and cytological changes in the MGs were more pronounced in patients with seasonal allergic conjunctivitis. Moreover, Wu et al. found that these relationships differed between adults and children (Wu et al., 2022a). Additionally, A 1-unit increase in PM2.5 levels is associated with a 0.06 increase in ocular inflammation and a 0.07 increase in gland dropout (Huang et al., 2020). This suggests that PM2.5 may damage the glands by inducing inflammation. However, most existing studies rely heavily on subjective assessment methods, such as manual analysis and qualitative classification, which compromise the reliability and reproducibility of their findings. These studies often fail to comprehensively capture the patterns of MG morphological changes across different populations.

With the advent of intelligent algorithms, the automated quantitative analysis of MG images has garnered increasing attention from researchers (Yeh et al., 2021; Xiao et al., 2021; Li et al., 2024; Yang et al., 2023; Dai et al., 2021). Deep learning-based segmentation methods for MG have been developed (Saha et al., 2022; Lin et al., 2024; Wu et al., 2024), allowing automated quantitative analysis of various MG parameters, such as length, width, and tortuosity, the latter referring to the twisting or winding nature of the glands, by learning features from large, unlabeled meibography datasets (Setu et al., 2021). AI-based quantitative analysis algorithms offer more objective and precise assessments of MG parameters, providing new tools and perspectives for studying changes in MGs associated with AC.

This study aims to utilize an intelligent quantitative analysis algorithm to systematically evaluate the microstructural changes of MGs in patients with AC, exploring the relationships between these changes and clinical relevant indicators, such as OSDI scores and TBUT. This study incorporates intelligent analysis technology to provide a more objective and accurate assessment of MGs than traditional methods to offer deeper insights into the diagnosis and management of AC.

Although previous studies have identified allergies as a potential risk factor for MG damage (Mizoguchi et al., 2017b; Chao and Tong, 2018; Wu et al., 2022b), these investigations have often been constrained by subjectivity and limited reproducibility. Recent advancements in AI have demonstrated high accuracy and reliability in evaluation of MG parameters (Xiao et al., 2021; Fasanella et al., 2016; Zhang et al., 2022). In this study, we utilized an AI-based quantitative analysis system to systematically examine the changes in MGs in patients with AC, aiming to explore the specific effects of AC on MGs and their correlation with clinically relevant parameters.

Our study found that compared to the control group, the AC group exhibited reduced tear meniscus height, shorter non-invasive first TBUT, increased conjunctival hyperemia, and significantly higher OSDI scores. The reduction in tear meniscus height may reflect decreased tear production or increased tear evaporation, leading to reduced tear retention. The shortened non-invasive first TBUT indicates decreased tear film stability, making the ocular surface more susceptible to external stimuli, causing discomfort. Additionally, the increased conjunctival hyperemia and elevated OSDI scores further confirm the exacerbation of ocular discomfort in AC patients. These findings are consistent with previous studies (Kim and Moon, 2013; Akasaki et al., 2022), which showed that dry eye symptoms are more severe in patients with AC, and the more allergens present, the worse the dry eye symptoms become.

In terms of MG morphology, the overall gland length in the AC group was shorter, and the central five glands showed significantly reduced length and area compared to the control group. This finding further supports the notion that AC affects MG structure, particularly in the central glands, which may be related to chronic inflammation of the ocular surface. The reduction in central gland length and area may result from glandular atrophy or dysfunction caused by local inflammatory responses. Studies have shown that chronic inflammation can trigger immune cells to release various inflammatory mediators, such as cytokines and chemokines, as well as bacterial imbalances, releasing endotoxins, leading to MG structure damage (Mizoguchi et al., 2017b). In our study, there were no significant differences in the width and tortuosity of the central 5 glands, which may be due to the fact that AC-induced inflammation is more likely to damage the openings or ends of the MGs and has less effect on the width and direction of growth of the MGs; and the direction of growth of the MGs is constrained by the anatomical structure of the eyelids, which also allows for the consistency of the overall tortuosity of the MGs. Additionally, mechanical stimulation of the eyes, such as frequent eye rubbing or changes in blinking patterns, may accelerate MG damage, causing glandular atrophy or dysfunction (Wang et al., 2018; Arita et al., 2010).

In the correlation analysis, the gland drop out ratio has a positive correlation of 0.19 with gender, suggesting that male patients may be more prone to gland dropout. This finding aligns with a recent meta-analysis (Hassanzadeh et al., 2021), which revealed that men are more likely to develop MGD, with an estimated overall prevalence of 35.8%. The reasons may be related to different lifestyle habits of men and women, such as smoking and drinking, which can cause some damage to the glands. In addition, it appears that androgens have an effect on gland shortening. The observed negative correlation of −0.24 between gland dropout ratio and OSDI contradicts previous findings, which may be explained by the fact that in allergic conjunctivitis, patient awareness is influenced by various factors, leading to a potential masking of the subjective perception of symptoms. Furthermore, the negative correlation between gland dropout ratio and age may indicate that as children grow older, their gland function gradually develops and matures, leading to a reduction in gland dropout.

Among patients with allergic conjunctivitis, gland count has a negative correlation of −0.21 with both the non-invasive first TBUT and non-invasive average TBUT, this suggests that an increased number of glands can lead to tear film damage in AC patients. The main reason to consider is still because, an increased number of glands may result in loss of epithelial cells and cup cells due to inflammation, produce abnormal or poor-quality lipids, imbalance the tear composition exacerbating tear film breakup (McCulley and Shine, 2004; Nelson et al., 2011). Gland area has a positive correlation of 0.17 with tear breakup time, which is consistent with previous research, suggesting that a more uniform and complete lipid layer helps to cover the tear film surface, reducing tear evaporation and maintaining tear film stability (Mizoguchi et al., 2017c). Conjunctival hyperemia has a positive correlation of 0.19 with OSDI, indicating that the inflammatory response in allergic conjunctivitis directly affects patients’ subjective symptoms, particularly worsening dry eye symptoms.

A key clinical value of our study lies in the use of AI algorithms to assess changes in the meibomian glands of AC patients, which enhances the precision, objectivity, and diversity of parameters. AI algorithms are capable of detecting even the smallest changes, which is especially useful in AC, where the meibomian glands may undergo varying degrees of inflammation or functional damage. Traditional methods often struggle to quantify these subtle pathological changes accurately. Furthermore, AI provides a quantitative evaluation of key meibomian gland parameters, such as gland length, width, area, counts and gland deformation. Additionally, by standardizing the evaluation process, AI algorithms significantly reduce human biases, resulting in more consistent and reliable assessments. These offers a powerful tool for tracking and monitoring the ocular surface conditions of AC patients during follow-up visits.

There are several limitations to this study. First, the analysis is restricted to parameters from the upper eyelid, as previous studies have shown that the upper eyelid provides clearer and higher-quality meibography images compared to the lower eyelid, minimizing uneven focus and demonstrating stronger correlations with clinical indicators (Deng et al., 2021; Daniel et al., 2019). Second, Further research is required to explore the distinct changes in the MGs across these different AC subtypes. Third, various environmental exposures could have also played a role in the results. Factors such as air pollution, climate conditions (e.g., humidity and temperature), and prolonged screen time are known to affect ocular health and meibomian gland function. Future studies should control for these confounding variables. Lastly, the cross-sectional design of this study limits the ability to establish causal relationships. Future longitudinal studies could provide more robust evidence by tracking changes in MG parameters and clinical symptoms over time in a cohort of AC patients.

This study systematically evaluated changes in the MGs of patients with AC using a novel quantitative analysis algorithm. The findings demonstrate that AC significantly affect the overall length of the MGs, as well as the length and area of the central five glands. These structural changes are strongly associated with decreased tear film stability, increased conjunctival hyperemia, and worsening ocular discomfort. These results underscore the detrimental impact of AC on MG function and overall ocular surface health.