Pre-diagnostic plasma endogenous steroids and the risk of exfoliation glaucoma

This nested case-control study found no statistically significant associations between pre-diagnostic plasma endogenous steroids and exfoliation glaucoma risk after multiple testing correction, though suggestive inverse trends for androgens and positive trends for progestogens warrant further investigation in larger studies.

The Gist

Imagine your eyes are like a high-tech city with a complex plumbing system. The "water" (fluid) needs to flow out smoothly to keep the pressure just right. In a condition called Exfoliation Glaucoma (XFG), a sticky, gummy substance (called exfoliation material) clogs the drains, causing pressure to build up and damaging the city's infrastructure (the optic nerve).

Scientists have long known that this "gum" clogs up more often in women than men, and that it seems to run in families. They also suspected that the body's internal chemical messengers—specifically steroid hormones (like testosterone, estrogen, and cortisol)—might be the architects behind this clogging.

This study was like a massive, high-stakes detective investigation to see if the levels of these hormones in a person's blood before they got sick could predict who would develop the eye disease.

The Investigation

The researchers acted like time travelers. They looked back at blood samples taken from over 400 people (217 who later developed the eye disease and 217 who didn't) from large, long-running health studies. These blood samples were taken roughly 12 years before the eye disease was diagnosed.

They tested the blood for 18 different steroid hormones. Think of these hormones as different types of "construction crews" in the body:

• Androgens (like testosterone): The "builders."
• Progestogens (like progesterone): The "regulators."
• Glucocorticoids (like cortisol): The "stress managers."
• Estrogens: The "female-specific architects."

What They Found

The team ran the numbers to see if having too much or too little of any specific "crew" meant you were more likely to get the clogged drains.

The Big Surprise:
Strictly speaking, no single hormone was a "smoking gun." After running the math to account for the fact that they were testing so many things at once, none of the hormones reached the level of "statistical certainty" usually required to say, "Aha! This hormone definitely causes the disease."

The "Almost" Clues:
However, the data did whisper some interesting hints that the researchers want to investigate further:

1. The "Anti-Clog" Crew (Androgens): There was a hint that men with higher levels of certain "builder" hormones (specifically a type called 11-ketotestosterone) might have been less likely to get the disease. It's as if having a stronger construction crew helped keep the drains clear.
2. The "Clog" Crew (Progestogens): Conversely, there was a hint that higher levels of progesterone might be linked to a higher risk. It's like having too many regulators accidentally ordering too much of the sticky gum.
3. The Stress Manager (Cortisone): Previous studies suggested that a stress hormone called cortisone might protect against the disease, and this study saw a similar (though not statistically proven) trend.

Why Didn't They Find a Definite Answer?

The researchers compared this to looking for a needle in a haystack. Because they tested 18 different needles at once, the rules of the game say you need a very strong signal to be sure you found the right one. The signals they found were "faint whispers" rather than "loud shouts."

Also, the study participants were mostly older adults. By the time you are older, your hormone levels have naturally changed and settled, which might have made it harder to see the original "blueprint" of how these hormones influence the disease decades earlier.

The Takeaway

Think of this study as a map that shows some interesting terrain but doesn't quite pinpoint the treasure yet.

• The Verdict: We can't say for sure that hormones cause Exfoliation Glaucoma based on this study alone.
• The Hope: The "whispers" of data suggest that androgens (male hormones) might be protective, while progesterone might be a risk factor. This gives scientists a new direction.

The Bottom Line:
The body's chemical balance is a complex orchestra. This study suggests that the "male" instruments might be playing a tune that keeps the eyes healthy, while the "female" instruments might be playing a tune that occasionally causes a jam. While we don't have the full sheet music yet, this research gives doctors and scientists a better idea of which instruments to listen to in future studies to prevent this sight-threatening condition.

Technical Summary

1. Problem Statement

Exfoliation glaucoma (XFG) is the most common form of secondary glaucoma, characterized by the accumulation of exfoliation material (XFM) in the eye, leading to elevated intraocular pressure (IOP) and optic nerve damage. While genetic factors (specifically LOXL1 variants) and environmental factors (age, UV exposure, latitude) are known risk factors, the underlying biological pathways remain incompletely understood.

Prior metabolomic studies suggested an inverse association between XFG and steroid-related plasma metabolites (particularly cortisone), and epidemiological data indicates higher incidence in women, suggesting potential links to reproductive history and sex hormones. However, previous studies were limited by:

• Evaluating relative metabolite levels rather than absolute concentrations.
• Lack of direct measurement of specific steroid hormones.
• Inability to assess absolute steroid concentrations roughly a decade prior to diagnosis.

This study aimed to elucidate the role of pre-diagnostic plasma endogenous steroid hormones in the etiology of XFG and XFG suspect (XFGS) by evaluating absolute concentrations of 18 specific steroids.

2. Methodology

Study Design and Population:

• Design: Nested 1:1 matched case-control study within three prospective US cohorts: Nurses' Health Study (NHS), NHSII, and Health Professionals Follow-up Study (HPFS).
• Participants:
  • Cases: 217 incident XFG/XFGS cases (confirmed via medical record review by a glaucoma expert).
  • Controls: 217 matched controls (no glaucoma diagnosis) selected via risk-set sampling.
  • Demographics: 62 men and 372 women (total N=434).
• Matching Factors: Age at blood draw, sex, ancestry, residential latitude/longitude, time/month/year of blood draw, years from blood draw to diagnosis, and menopausal/hormone therapy status (for women).
• Follow-up: Blood samples were collected 1989–1999 (average 11.8 years prior to diagnosis).

Exposure Measurement:

• Analytes: 18 endogenous steroids measured in plasma using Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS).
• Steroid Classes:
  • Androgens: Testosterone, Androstenedione, DHEAS, 11-ketotestosterone, 11β-hydroxyandrostenedione.
  • Estrogens: Estradiol, Estrone, Estrone sulfate.
  • Progestogens: Progesterone, 17α-hydroxyprogesterone, Pregnenolone sulfate.
  • Mineralocorticoids: Aldosterone, Deoxycorticosterone, Corticosterone, 18-hydroxycorticosterone.
  • Glucocorticoids: Cortisol, Cortisone, 11-deoxycortisol.
• Quality Control: High reproducibility (CV ≤15%, ICC ≥0.75 for most steroids) confirmed via pilot studies.

Statistical Analysis:

• Primary Model: Multivariable-adjusted conditional logistic regression.
  • Model 1: Adjusted for matching factors.
  • Model 2: Model 1 + lifestyle/clinical covariates (smoking, BMI, physical activity, sunlight exposure, diet, comorbidities, oral steroid use).
• Multiple Testing Correction:
  • Individual Steroids: Number of Effective Tests (NEF) approach (Significance threshold: NEF < 0.05; Worthy of investigation: NEF < 0.2).
  • Steroid Classes: False Discovery Rate (FDR) (Significance threshold: FDR < 0.05).
• Secondary Analyses: Effect modification by age, residential latitude, and disease severity (XFG with visual field loss vs. XFGS without). Metabolite Set Enrichment Analysis (MSEA) was used for class-level associations.

3. Key Results

Overall Associations:

• No Statistically Significant Findings: After correction for multiple testing (NEF and FDR), no individual steroid or steroid class showed a statistically significant association with XFG/XFGS risk in the overall population or in sex-specific strata.
• Suggestive Trends (Nominal Significance):
  • Androgens: Showed a non-significant inverse trend (FDR=0.22). 11-ketotestosterone showed a nominal inverse association (OR=0.54; 95% CI: 0.31–0.93; P=0.03).
  • Progestogens: Showed a positive trend (FDR=0.31). Progesterone showed a borderline positive association (OR=2.21; 95% CI: 1.00–4.87; P=0.05).
  • Cortisone: Showed an inverse association (OR=0.36; 95% CI: 0.13–1.03), consistent with prior metabolomic data but not reaching nominal significance.

Sex-Specific Findings:

• Females: Progesterone and deoxycorticosterone were significantly positively associated with risk in Model 2.
• Males: Nine endogenous steroids (including 11-ketotestosterone, testosterone, and various corticosteroids) showed nominally significant inverse associations with risk.
• Interactions: Significant interactions by sex were observed for androstenedione, progesterone, 17α-hydroxyprogesterone, and several mineralocorticoids (p < 0.05).

Disease Severity (Heterogeneity):

• Analyses separating XFG (with visual field loss) from XFGS (without visual field loss) revealed that the adverse association with progestogens was FDR-significant for XFG (severe) but not XFGS.
• A suggestive inverse trend for the androgen class was observed specifically in XFG cases, driven largely by 11-ketotestosterone.

4. Key Contributions

1. Absolute Quantification: Unlike previous metabolomic studies that measured relative ratios, this study utilized LC-MS/MS to measure absolute concentrations of 18 specific steroids, providing a more direct assessment of steroid pathways.
2. Longitudinal Pre-diagnostic Data: Samples were collected an average of 11.8 years prior to diagnosis, minimizing reverse causation bias (i.e., the disease altering hormone levels).
3. Comprehensive Steroid Panel: The study covered a broad spectrum of steroid classes (androgens, estrogens, progestogens, mineralocorticoids, glucocorticoids) rather than focusing on a single hormone.
4. Sex-Specific Nuance: The study highlighted distinct sex-specific patterns, particularly the protective inverse associations in men versus positive associations in women for certain steroids.

5. Significance and Conclusion

Significance:
While the study did not yield statistically significant associations after strict multiple testing correction, the suggestive patterns provide biological plausibility for steroid involvement in XFG pathogenesis:

• Androgens: The inverse trend (particularly 11-ketotestosterone) supports the hypothesis that androgens may downregulate TGF-β1 signaling, a pathway implicated in the fibrotic processes of XFG.
• Progestogens/Estrogens: The positive trend with progesterone aligns with previous findings linking higher lifetime estrogen/progesterone exposure to increased XFG risk, potentially via upregulation of LOXL1 and TGF-β1.
• Cortisone: The inverse trend reinforces prior metabolomic findings, suggesting a potential link between glucocorticoid metabolism and XFG, possibly mediated through glucose metabolism and glycation of basement membranes.

Limitations:

• Statistical Power: The sample size (217 cases) may be underpowered to detect modest associations after rigorous multiple testing corrections.
• Cohort Homogeneity: The population consists primarily of highly educated White health professionals, limiting generalizability to other ancestral groups.
• Sample Storage: Long-term storage of blood samples (up to 30 years) raises theoretical concerns about biomarker degradation, though quality control measures supported stability.

Conclusion:
This study supports the hypothesis that endogenous steroid pathways play a role in XFG etiology. Although no single association reached statistical significance, the observed trends—specifically the potential protective role of androgens and the risk-associated role of progestogens—warrant further investigation in larger, more diverse prospective studies to confirm these pathways and potentially identify therapeutic targets.