Sex hormone-mediated corneal remodeling in keratoconus: a multivariate analysis of tomographic and biomechanical variations across gender, age, and menstrual phases.

This cross-sectional study was conducted at the Peking University Third Hospital Eye Center from June 2020 to March 2022, with approval granted by the local ethics committee in accordance with Declaration of Helsinki guidelines. Informed consent was obtained from all individual participants or, where participants were under 16, from their parent(s) or legal guardian(s). The study enrolled 180 KC patients (KC group) and controls with mild-to-moderate myopia undergoing laser vision correction (LVC group) from the Peking University Third Hospital Eye Center, with one eye per participant randomly selected for analysis. Diagnostic criteria required: (1) Positive clinical history; (2) Stage I-IV on Amsler-Krumeich classification; (3) Presence of ≥ 1 characteristic slit-lamp findings (Fleischer ring, Vogt striae, or Munson sign); confirmed by two independent corneal specialists. Exclusion criteria were: (1) Active ocular disease (e.g., dry eye syndrome, glaucoma); (2) History of ocular surgery or trauma; (3) Current hormone therapy; (4) Systemic disorders affecting connective tissue (e.g., Marfan syndrome, Ehlers-Danlos syndrome). Contact lens discontinuation periods were ≥ 2 weeks for soft lenses and ≥ 4 weeks for rigid gas permeable lenses. All patients were divided into three age groups: ≤ 20 years, 21–30 years, and > 30 years. Female patients were further categorized into three groups based on their menstrual cycle: follicular phase, ovulatory phase, and luteal phase. All patients received a complete standard of ophthalmological examination, including visual acuity, slit-lamp, fundus, corneal morphological, and biomechanical examinations. Corneal tomography was performed using Pentacam HR (Oculus Optikgeräte GmbH, Wetzlar, Germany), while biomechanical assessment utilized Corvis ST II (Oculus Optikgeräte GmbH, Wetzlar, Germany). All the Pentacam HR and Corvis ST II examinations were performed by the same skilled technician in the same examining room under the same light conditions between 8: 00 and 17: 00 to avoid bias. Sixteen tomographic parameters were analyzed including: central corneal thickness (CCT), thinnest corneal thickness (TCT), maximum simulated keratometry (K max ), index of surface variance (ISV), index of vertical asymmetry (IVA), keratoconus index (KI), center keratoconus index (CKI), index of height asymmetry (IHA), index of height decentration (IHD), minimum sagittal curvature (Rmin), deviation of front elevation difference map (Df), deviation of back elevation difference map (Db), deviation of average pachymetric progression index (Dp), deviation of minimum thickness (Dt), deviation of Ambrósio’s relational thickness maximum (Da), and Belin/Ambrosio enhanced ectasia total deviation index (D). Nine corneal biomechanical parameters were analyzed including: stiffness parameter at first applanation (SP-A1), Integrated Radius (IR), Ambrósio relational thickness in the horizontal profile (ARTh), the ratio between DA at the apex and the average of the DAs at two mm around the centre in the temporal and nasal directions (DA ratio), the inferior-superior difference keratometry (IS value), Pentacam HR random forest index (PRFI), Corvis biomechanical index (CBI), and tomographic and biomechanical index (TBI). Venipuncture was standardized between 10:00 and 13:00 to control for diurnal hormone variation, with serum aliquots stored at -80 °C until analysis. Serum T, E 2 , and P levels were measured using chemiluminescent microparticle immunoassay. Simultaneously, information regarding the menstrual cycle of female patients was collected [ 12 ]. All analyses were conducted in SPSS, Version 26 (IBM Corp., Armonk, NY). The Kolmogorov-Smirnov test was used to test for normality. All the quantitative data was described as mean ± SD. The Mann-Whitney U-test was used to evaluate differences in parameters between KC and LVC group. Spearman correlation analysis was used to explore the correlations between sex hormones and different corneal parameters. A P  < 0.05 was considered statistically significant. A Bonferroni correction was applied to adjust the significance level for planned pairwise comparisons between groups. For the exploratory correlation analyses involving multiple parameters, the false discovery rate (FDR) was controlled instead, using the Benjamini-Hochberg method. Only associations with an FDR-adjusted (q value) < 0.05 were considered significant in the correlation analyses.

The final cohort comprised 435 subjects (KC: n  = 180; LVC: n  = 255) with complete datasets. Table  1 shows the demographic and serum sex hormones data of each group. Mean age did not differ significantly between the KC group and the LVC group, regardless of whether the patients were female (KC: 26.18 ± 5.77 years vs. LVC: 27.30 ± 6.26 years; P  = 0.238) or male (KC: 24.41 ± 6.14 years vs. LVC: 24.71 ± 7.14 years; P  = 0.582). KC patients showed significantly lower T levels versus controls (males: 11.73 ± 4.17 vs. 13.17 ± 4.57 nmol/L, P  = 0.013; females: 0.84 ± 0.34 vs. 0.96 ± 0.54 nmol/L, P  = 0.041). Male KC patients exhibited lower E 2 levels compared to controls (95.23 ± 27.80 vs. 110.60 ± 28.25 pmol/L; P  < 0.001). Table 1 Demographics parameters of study subjects in the KC and LVC groups KC LVC P value All Gender(female/male) 180(51/129) 255(115/140) - Female Age(years) 26.18 ± 5.77 27.30 ± 6.26 0.238 T (nmol/L) 0.84 ± 0.34 0.96 ± 0.54 0.041 E 2 (pmol/L) 335.74 ± 281.59 351.32 ± 340.73 0.967 P (nmol/L) 7.33 ± 10.60 7.83 ± 12.42 0.349 CCT 465.31 ± 38.52 544.37 ± 27.59 < 0.001 TCT 440.47 ± 40.90 539.87 ± 27.45 < 0.001 K max 59.67 ± 7.54 44.67 ± 1.63 < 0.001 D 10.89 ± 4.17 1.19 ± 0.52 < 0.001 CBI 0.93 ± 0.21 0.08 ± 0.14 < 0.001 TBI 0.99 ± 0.06 0.17 ± 0.14 < 0.001 Male Age(years) 24.41 ± 6.14 24.71 ± 7.14 0.582 T (nmol/L) 11.73 ± 4.17 13.17 ± 4.57 0.013 E 2 (pmol/L) 95.23 ± 27.80 110.60 ± 28.25 < 0.001 P (nmol/L) 1.43 ± 0.50 1.48 ± 0.49 0.342 CCT 478.93 ± 44.01 538.31 ± 36.43 < 0.001 TCT 463.54 ± 45.97 533.43 ± 37.92 < 0.001 K max 56.95 ± 9.16 45.33 ± 4.39 < 0.001 D 8.80 ± 4.97 1.61 ± 2.47 < 0.001 CBI 0.83 ± 0.31 0.11 ± 0.29 < 0.001 TBI 0.93 ± 0.22 0.28 ± 0.30 < 0.001 KC, keratoconus; LVC, laser vision correction; T, testosterone; E 2 , estradiol; P, progesterone; CCT, central corneal thickness; TCT, thinnest corneal thickness; K max , maximum simulated keratometry; D, Belin/Ambrosio enhanced ectasia total deviation index; CBI, Corvis biomechanical index; TBI, tomographic and biomechanical index.Mann-Whitney U-test; all data are expressed as mean ± SD. Demographics parameters of study subjects in the KC and LVC groups KC, keratoconus; LVC, laser vision correction; T, testosterone; E 2 , estradiol; P, progesterone; CCT, central corneal thickness; TCT, thinnest corneal thickness; K max , maximum simulated keratometry; D, Belin/Ambrosio enhanced ectasia total deviation index; CBI, Corvis biomechanical index; TBI, tomographic and biomechanical index.Mann-Whitney U-test; all data are expressed as mean ± SD. A total of 269 males were enrolled and stratified into three age groups: 94 individuals in the ≤ 20 years group, 129 in the 21–30 years group, and 46 in the > 30 years group. As shown in Fig.  1 , E 2 levels were significantly lower in the male KC group compared to the male LVC group across all age groups ( P  < 0.017 after Bonferroni correction). T levels between the KC and LVC groups differed most markedly in the 21–30 years group ( P  = 0.031). P levels showed no significant differences between the KC and LVC groups in any of the three male age groups. Fig. 1 Male serum testosterone (T), estradiol (E 2 ), and progesterone (P) levels were analyzed and compared between control (LVC) and keratoconus (KC) groups across three different age groups (≤ 20, 21–30, > 30). The Mann-Whitney U test was used to assess group differences, * P  < 0.05, ** P  < 0.01 Male serum testosterone (T), estradiol (E 2 ), and progesterone (P) levels were analyzed and compared between control (LVC) and keratoconus (KC) groups across three different age groups (≤ 20, 21–30, > 30). The Mann-Whitney U test was used to assess group differences, * P  < 0.05, ** P  < 0.01 Sex hormone levels were compared between the female KC group and the female LVC group, stratified by menstrual phases and age. A total of 166 female participants were stratified by menstrual cycle phase: 82 in the luteal phase, 81 in the follicular phase, and 3 in the ovulatory phase. In the luteal phase, T levels were significantly lower in the KC group than in the LVC group ( P  = 0.031; Fig.  2 ), with the most pronounced difference observed in the 21–30 years group ( P  = 0.045; Fig.  3 ). No significant intergroup differences were found for E 2 or P levels in any of the three hormones in the follicular phase. Fig. 2 Serum testosterone (T), estradiol (E 2 ), and progesterone (P) levels in females during the follicular and luteal phases were analyzed and compared between control (LVC) and keratoconus (KC) groups. The Mann-Whitney U test was used to assess group differences, * P  < 0.05 Serum testosterone (T), estradiol (E 2 ), and progesterone (P) levels in females during the follicular and luteal phases were analyzed and compared between control (LVC) and keratoconus (KC) groups. The Mann-Whitney U test was used to assess group differences, * P  < 0.05 Fig. 3 Serum testosterone (T), estradiol (E 2 ), and progesterone (P) levels were analyzed and compared between control (LVC) and keratoconus (KC) groups among females in the luteal phase across three age groups (≤ 20, 21–30, > 30). The Mann-Whitney U test was used to assess group differences, * P  < 0.05 Serum testosterone (T), estradiol (E 2 ), and progesterone (P) levels were analyzed and compared between control (LVC) and keratoconus (KC) groups among females in the luteal phase across three age groups (≤ 20, 21–30, > 30). The Mann-Whitney U test was used to assess group differences, * P  < 0.05 Serum E 2 levels in male patients ≤ 20 years old were positively correlated with Rmin ( r  = 0.301, q = 0.012), and negatively correlated with Df ( r = -0.357, q < 0.001), CKI ( r = -0.343, q = 0.012), ISV ( r = -0.312, q = 0.012), KI ( r = -0.312, q = 0.012), K max ( r = -0.301, q = 0.012), Db, IR, IS Value, D, CBI, and TBI(Fig.  4 (a)). Serum E 2 levels in male patients > 30 years old was positively correlated with ARTh ( r  = 0.408, q = 0.042), and negatively correlated with IHA ( r = -0.447, q = 0.042), IR ( r = -0.397, q = 0.042), Dp ( r = -0.393, q = 0.042), IHD ( r = -0.372, q = 0.048), Da ( r = -0.362, q = 0.048), and D ( r = -0.360, q = 0.048) (Fig.  4 (b)). Serum T levels in male patients aged 21–30 were negatively correlated with PRFI, IHA, IS Value, K max , and Da ratio (Fig.  4 (c)). Fig. 4 Spearman correlation analysis of serum hormone levels with corneal parameters in male patients across age groups. ( a ) Correlations with serum estradiol (E 2 ) levels in patients ≤ 20 years old. ( b ) Correlations with serum E 2 levels in patients > 30 years old. ( c ) Correlations with serum testosterone (T) levels in patients aged 21–30. Statistical significance was defined as an FDR-adjusted (q value) < 0.05 Spearman correlation analysis of serum hormone levels with corneal parameters in male patients across age groups. ( a ) Correlations with serum estradiol (E 2 ) levels in patients ≤ 20 years old. ( b ) Correlations with serum E 2 levels in patients > 30 years old. ( c ) Correlations with serum testosterone (T) levels in patients aged 21–30. Statistical significance was defined as an FDR-adjusted (q value) < 0.05

While KC pathogenesis remains multifactorial, current evidence implicates genetic predisposition, immune dysregulation, and metabolic abnormalities in disease progression. Emerging evidence suggests that endocrine pathways are associated with the pathophysiology of KC. Steroid hormones exert pleiotropic effects via systemic circulation, with ocular tissues expressing functional sex hormone receptors that regulate corneal homeostasis [ 13 ]. Bassiouny found that patients with hyperthyroidism tended to have abnormal tomographic parameters correlating with KC [ 14 ]. Lenk tested the hair cortisol concentration and found that patients with progressive KC consistently exhibited higher cortisol concentrations [ 15 ]. Regarding sex hormones, many studies have shown that they contribute to KC. Escandon et al. found that gonadotropin-releasing hormone level was lower in KC patients. Through vitro experiments, they further demonstrated that gonadotropin-releasing hormone could modulate hormonal and fibrotic responses in the KC corneal stroma [ 16 ]. Karamichos first found that the LH/FSH ratio in KC patients was significantly lower than that in healthy controls [ 17 ]. Emilio et al. documented accelerated keratoconus progression in a 49-year-old female following tibolone therapy for endometriosis [ 18 ]. Yuksel et al. found that women treated with in-vitro fertilization consistently experienced progression in KC [ 19 ]. Hormonal changes resulting in alterations in corneal biomechanics were considered a significant factor in the progression of KC during pregnancy [ 20 , 21 ]. However, such isolated cases cannot establish causality and primarily underscore the need for systematic investigation into the role of sex hormones. Estrogen receptor, androgen receptor, and progesterone receptor have been confirmed to be expressed in corneal epithelial, stromal, and endothelial cells [ 22 ]. Ayan et al. demonstrated that the mRNA expressions of estrogen receptor α and androgen receptor were significantly higher in patients with keratoconus [ 23 ]. However, some studies have demonstrated that sex hormones, such as androgens, may have anti-inflammatory effects [ 24 ]. The effect of estrogen on inflammation is context-dependent, modulating immune responses to exert both anti-inflammatory and, in certain settings, pro-inflammatory effects. Suzuki et al. found that 17-β-estradiol increased the expression of pro-inflammatory genes in cultured corneal epithelial cells, indicating significant differences in the functional regulation of hormones and their impact on the eye [ 25 ]. There were also other studies finding that 17-β-estradiol treatment significantly reduced the matrix metalloproteinase-2 mRNA in human corneal stromal cells as well as the matrix metalloproteinase-2 proteins [ 26 ]. Therefore, when evaluating the impact of sex hormones on keratoconus progression, it is critical to account for the heterogeneous effects among different hormones, the dual roles of individual hormones, and their intricate network of interactions. Considering the phenomenon that KC often progresses during puberty and pregnancy, and stabilizes after menopause, we decided to investigate the difference in sex hormones between patients with KC and the LVC subjects. Pentacam HR is the most widely used corneal morphological examination device and it provides detailed measurements of corneal curvature and anterior and posterior surface elevation [ 27 ]. It is well known that the progression of KC leads to the destruction of corneal stroma, resulting in corneal biomechanics and weakened mechanical strength [ 28 ]. With the recent development of corneal biomechanics in vivo, more and more studies have focused on the biomechanical parameters of cornea obtained. The assessment of corneal biomechanics has been successfully used to diagnose KC and evaluate the surgical outcomes. In this study, Corvis ST II was used to assess biomechanical parameters of eyes KC and LVC group. To analyze the effects of sex hormones on KC, the Spearman’s correlation coefficient was used to investigate the association between E 2 and T levels and the morphology and biomechanics of KC. The Mann-Whitney U-test was used to evaluate differences in parameters between the KC and LVC groups. Our cohort analysis revealed persistently lower E 2 levels across all male age strata in the KC group compared with controls. T levels were significantly different in females aged 21–30 in the luteal phase and in male patients aged 21–30 years. These findings contrast with our early pilot study (KC: n = 62; LVC: n = 120), underscoring the importance of adequate statistical power [ 8 ]. When analyzed by menstrual cycle phase and age, E 2 and T levels showed negative correlations with most corneal morphological parameters. Furthermore, E 2 and T levels positively correlated with Rmin, ARTh, and SP-A1, and negatively with CBI, TBI, and IR. These patterns suggest an association between higher levels of these two hormones and more favorable corneal biomechanical and morphological profiles. We hypothesize that the generally lower E 2 level in males, and an age-related decrease in T in younger males, may partially contribute to the observed higher incidence of KC in these groups. While epidemiological evidence links sex steroids to KC risk, their precise pathophysiological role remains complex and potentially context-dependent [ 29 ]. By leveraging a large sample, our study delineates variations in sex hormones across genders, ages, and menstrual cycles, and their associations with corneal metrics. We found that in male KC patients, higher E 2 levels were linked to more favorable parameters across all ages, whereas higher T levels showed this association primarily in the 21–30-year group. In female patients, higher T levels were associated with better parameters in the 21–30-year luteal-phase group. The potential role of E 2 and T in other ocular diseases has been explored. For example, a study found that corneal nerve regeneration is faster in female mice than in male mice, suggesting that E 2 plays an important role in corneal nerve regeneration [ 30 ]. Other studies have shown that estrogen may improve certain diseases by inhibiting the NLRP3 inflammasome [ 31 ]. Furthermore, sex hormones are generally recognized for their ability to mitigate ocular surface inflammation [ 13 , 32 , 33 ]. Therefore, we speculate that sex hormones, such as E 2 and T, may influence KC through pathways that potentially modulate corneal inflammation. However, this proposed mechanism requires validation through longitudinal and experimental studies. To our knowledge, this study represents the first multidimensional analysis integrating age stratification, menstrual phase, and biomechanical profiling in KC research. This study has several limitations. This study has several limitations. First, its cross-sectional design precludes the establishment of causal relationships between sex hormones and corneal parameters. Second, hormone levels were measured at a single time point, which may not accurately reflect an individual’s long-term hormonal milieu given physiological fluctuations. Third, our control group comprised individuals scheduled for LVC, who may not represent a hormonally “normal” population free of ocular conditions, potentially limiting generalizability. Fourth, the sample size of the ovulatory phase group was too small for robust statistical analysis. Finally, the exploratory correlation analyses involved numerous comparisons without formal statistical correction, increasing the risk of Type I errors. These findings should be interpreted within the context of these constraints, and they highlight the need for longitudinal studies with repeated hormone assessments, objectively measured menstrual cycles, and carefully matched controls to confirm and extend our observations.

In conclusion, our study reveals distinct associations between sex hormones and KC, with higher E 2 levels consistently linked to more favorable corneal parameters across all ages in males, and luteal-phase T levels showing specific associations in females. The 21–30-year group demonstrated the strongest hormone-cornea associations, highlighting a period of particular interest for further research. In particular, monitoring hormone levels and exploring hormonal modulation strategies in KC patients aged 21–30 warrants future research. These findings position sex hormones as a relevant factor in understanding KC heterogeneity. Based on existing evidence of their anti-inflammatory properties, we hypothesize that sex hormones might influence KC pathophysiology through immunomodulatory pathways, a mechanism that warrants direct investigation in future studies. Ultimately, clarifying these relationships may inform future strategies for KC management.

Keratoconus (KC) is a progressive bilateral corneal ectatic disorder characterized by corneal thinning, irregular astigmatism, and subsequent vision impairment [ 1 ]. Typically manifesting during adolescence, KC progression generally stabilizes by the third decade, with reported prevalence rates of 0.004%-0.06% and annual incidence rates of 0.05%-0.23% [ 2 , 3 ]. Males exhibit earlier disease onset, with a documented sex ratio (male: female) ranging from 0.9:1 to 2.5:1 [ 4 ]. Adolescent and young adult KC patients often develop clinically significant, irreversible visual morbidity. The multifactorial pathogenesis involves genetic predisposition, biomechanical stress (e.g., eye rubbing), environmental exposures (including contact lens use and air pollution), and oxidative-inflammatory pathways [ 5 – 7 ]. Nevertheless, the precise molecular mechanisms underlying KC pathogenesis remain elusive, highlighting the need for continued investigation. Emerging, though still inconclusive, evidence suggests that sex hormones—particularly estradiol (E 2 ), progesterone (P), and testosterone (T)—may influence KC progression [ 8 – 11 ]. But few studies have investigated how sex hormones affect the cornea in patients with KC. This study aims to explore the complex and incompletely understood associations between sex hormones and corneal pathophysiology by comprehensively evaluating serum hormone profiles alongside morphological and biomechanical corneal parameters.