This post exists because both "it's a hoax" and "it's the master biomarker" get the science wrong. People inside the mold and CIRS community know contrast sensitivity testing as part of a familiar protocol. People inside mainstream clinical research often haven't heard of it, and when they do, they regard it skeptically. The two groups talk past each other, often heatedly, and the patient in the middle has no clear way to read the evidence.
We built a free, calibrated contrast sensitivity test, and many of the people who land on it arrive from CIRS communities. The honest thing to do is lay out what is documented, what is contested, and what neither side can fairly claim is settled. We will not pretend the contested science is closed in either direction.
Why this post exists
Patient communities organised around mold exposure and chronic inflammatory response syndrome (CIRS) frequently cite a visual contrast sensitivity (VCS) test as a screening tool. The framework comes largely from the clinical work of Dr. Ritchie Shoemaker, developed in his practice and published through patient-protocol literature across the 1990s and 2010s. Within those communities, VCS is often the first measurement that registered an objective change when other tests came back unremarkable, and that experience carries weight.
The framework is also contested. Some clinicians and methodology researchers consider parts of it scientifically underdeveloped — particularly the claim that contrast sensitivity in a specific spatial-frequency band is a specific signal for biotoxin exposure. The disagreement is not whether contrast sensitivity is a real visual measurement (it is, with decades of clinical use). It is over what a reduced result in a particular band actually tells you.
Our position is narrow: a contrast sensitivity test measures contrast sensitivity. Whether reduced sensitivity in any band is a specific signal for any specific etiology is a question the test itself cannot answer. The rest of this post lays out what's published, what's claimed, and where the disagreement lives.
Background: what the Shoemaker framework claims
Within the CIRS framework as typically described in patient-facing materials, exposure to water-damaged buildings — and to the mycotoxins, endotoxins, beta-glucans, and other inflammagens found in those buildings — can trigger a chronic, systemic inflammatory response in genetically susceptible individuals (often described in terms of certain HLA-DR haplotypes). The proposed clinical picture includes a multi-system symptom cluster (fatigue, cognitive complaints, headaches, mood changes, autonomic symptoms) along with characteristic findings on a panel of inflammatory and neuroendocrine labs.
One of the early and sensitive findings in the framework is reduced visual contrast sensitivity, specifically in the mid-range spatial frequencies — usually identified as 6 and 12 cycles per degree (cpd). The framework holds that biotoxins impair the optic nerve's processing of these frequencies within roughly 24 to 36 hours of exposure, and that those frequencies recover as the patient progresses through treatment. On that account, VCS becomes both a screening signal at intake and a way to track environmental change and treatment response over time.
The standard implementation is the Functional Acuity Contrast Test (FACT) chart, presented either as a physical chart at a fixed viewing distance or as a digital reproduction of it. The chart shows five rows of sine-wave grating patches at five spatial frequencies (1.5, 3, 6, 12, and 18 cpd), with nine contrast steps per row. The observer reports the orientation of each patch (left tilt, right tilt, or vertical) until they can no longer reliably identify it; the last correctly identified patch in each row defines the threshold for that frequency.
A note on scope. The CIRS framework is a clinical protocol — diagnostic criteria, a battery of labs, a sequenced treatment plan — developed by a single clinician across decades of practice and elaborated in patient-protocol books and clinic literature. When this post talks about "the Shoemaker framework," we mean that body of clinical literature, summarised neutrally. We are not endorsing its specific diagnostic claims, and we are not dismissing them.
What the broader literature says
The independent peer-reviewed literature on contrast sensitivity in biotoxin-exposed or mold-exposed populations specifically is sparse compared to better-studied conditions. There is a substantial CSF literature in cataract, glaucoma, multiple sclerosis, diabetic retinopathy, and traumatic brain injury — replicated across many labs and decades. The literature on CSF in environmental medicine is small, mostly internal to the framework's own clinic literature, and has not been independently replicated at the same scale.
That gap cuts both ways. It means the CIRS-pattern claim has not been independently confirmed in adequately blinded studies on general populations. It also means it has not been independently refuted — absence of high-quality replication is absence of evidence, not evidence of absence. The honest description: the within-clinic clinical observation is consistent across the framework's own literature; independent peer-reviewed replication of the specific 6-and-12-cpd specificity claim has not been published at the scale required to settle the question.
A few things are well established by the broader CSF literature, and bear on the question even though they were not done in biotoxin-exposed cohorts:
- Contrast sensitivity is a real, sensitive functional measure of visual processing. It samples a richer range of visual function than acuity, and can be reduced before acuity drops. Settled across decades of independent work (the Pelli-Robson chart, established in Pelli, Robson & Wilkins, 1988, is one anchor in that literature).
- Mid-frequency CSF loss is common across many conditions. Glaucoma (early), cataract, multiple sclerosis, diabetic retinopathy, post-concussion vision changes, refractive error, dry eye, and normal aging can all reduce contrast sensitivity in the mid-frequency range. This is not specific to any one cause.
- Test-retest reliability for the FACT chart specifically is a known concern. Reviews of the methodology — most directly Pelli & Bex (2013) in Vision Research — note that grating charts with discrete contrast steps have markedly worse reliability than letter charts (Pelli-Robson) or adaptive procedures (qCSF), particularly at low and middle frequencies. Independent cohorts have reported large fractions of healthy young adults scoring the maximum possible value at 1.5, 3, and 6 cpd on the FACT chart — sometimes more than 80% of healthy adults at 3 cpd — a known ceiling effect that limits the chart's ability to detect small changes.
- Major professional bodies have engaged with mold-related illness without endorsing VCS as a validated diagnostic instrument for it. The 2004 National Academies (Institute of Medicine) report Damp Indoor Spaces and Health is the most-cited reference here; it reviewed the literature on health effects of damp indoor environments and reached conclusions the CIRS community considers conservative and that mainstream public-health bodies consider appropriately cautious. We will not paraphrase its specific conclusions about VCS here — that is a primary source worth reading directly if it bears on your decisions, rather than filtered through a blog post.
A separate honest note: the framework's reported in-clinic accuracy figures for VCS (often cited as 92–98%) come from patient populations selected for multi-symptom presentation in his specialty practice. Selected-population accuracy is a different statistical animal from general-population sensitivity and specificity. The question independent researchers raise is whether the same figures hold up in adequately blinded designs. That replication has not been published.
Where the disagreement actually lives
Strip away the polemics and there are three substantive disagreements worth naming.
Specificity. The strongest version of the framework's CSF claim is that the 6-and-12-cpd reduction is specific enough to biotoxin exposure to be useful as a screening signal for it. The strongest version of the mainstream critique is that mid-frequency CSF loss has too many other possible causes — cataract, glaucoma, refractive error, fatigue, post-concussion changes, aging, an uncalibrated screen — to carry that specificity. Both are stated more cautiously by their careful proponents than by their loudest ones. The careful CIRS claim is that VCS is one input alongside HLA-DR genetics, symptom clusters, environmental history, and labs — not a standalone test. The careful mainstream critique is that nonspecific signals can still be useful in the right context.
Methodology. The FACT chart's nine discrete contrast steps and ceiling effects at low-to-mid frequencies mean its resolution — the smallest change it can register — is one full step, roughly 0.15 log units. Modern adaptive procedures — 2-down-1-up staircases (Levitt 1971), QUEST (Watson & Pelli, 1983), qCSF (Lesmes, Lu, Baek & Albright, 2010) — give threshold estimates roughly 3× more precise. A test designed to track recovery over weeks of remediation arguably needs the finer-resolution methodology. This is a methodology critique, not a critique of the underlying framework: the framework's clinical observations could be entirely correct and the chart could still be a coarse instrument for measuring them.
Clinical decision-making. Even where the framework's clinical observations are accepted on their own terms, what a CSF reading tells a clinician depends on what they are trying to decide. Sensitivity is not specificity. A test that turns up reduced CSF in many patients with documented biotoxin exposure is a screening signal, useful for raising or lowering the prior for further workup. A test that distinguishes biotoxin-exposed patients from those with cataract, glaucoma, or post-concussion change would be a specificity test, and would require a different study design to establish. Most of what the literature documents about VCS in this space is the sensitivity end of that distinction. The specificity end is where independent replication has not been published.
The honest framing: the Shoemaker clinical observation — within-patient CSF changes that track environmental and treatment events in his clinic population — has not been published in the form that would close the question in either direction. That doesn't make it false. It does make it not yet proven.
What we (VCS-Test) think
We are a free, well-calibrated, modern adaptive contrast sensitivity test, built deliberately not as a FACT clone. Our methodology is documented at /methodology; briefly, we use a 2-down-1-up adaptive staircase per spatial frequency, with display gamma correction, viewing-distance estimation, and credit-card screen-size calibration on a mid-gray surround. The output is a log-CS-versus-cpd curve across the frequencies your screen and viewing distance can faithfully reproduce.
If you suspect environmental illness, a contrast sensitivity reading is one data point. It is not the only one. A thoughtful intake includes environmental history (where you live, where you work, building age, recent water events, ventilation), symptom journaling over weeks not days, an eye exam to rule out refractive and ocular sources of contrast loss, and — depending on the workup pathway — a panel of inflammatory and neuroendocrine labs that the CIRS clinical literature describes. Outside that literature, IgE testing for specific mold allergens, ERMI (Environmental Relative Moldiness Index) dust sampling, and HERTSMI-2 scoring are the environmental tests CIRS practitioners typically run; whether those are the right tests for you is a question for the clinicians you work with.
We will not tell you our test is a "biotoxin screen." It isn't, and no contrast sensitivity test we know of meets the bar to be called one. We also will not tell you the measurement is worthless if your community values it. It is a real measurement of a real visual function, and the within-person change over time on a well-calibrated setup is a useful piece of objective data. Use the result as: (a) a baseline you can track over time, (b) one input to bring to clinicians familiar with your situation, (c) a screening signal — not a diagnosis of anything.
What it can and can't tell you
Note. This is a screening and tracking measurement, not a diagnosis.
A contrast sensitivity test does not diagnose CIRS, mold illness, biotoxin exposure, or any other specific condition. No CSF test does. The framework that incorporates CSF as one input describes a multi-element clinical protocol; the test on its own is one of many inputs.
Reduced contrast sensitivity in any band has many possible causes. Cataract, glaucoma (early), multiple sclerosis, optic neuritis, diabetic retinopathy, post-concussion vision changes, uncorrected refractive error, dry eye, fatigue, certain medications, and normal aging can all reduce contrast sensitivity, including in the mid-frequency band sometimes singled out in CIRS discussions. A reduction is consistent with the framework's pattern; it is not specific to it.
A single result is a snapshot. Test-retest variation is real even with clinical instruments — the Pelli-Robson chart has a test-retest repeatability of about ±0.15 log units, and the smallest clinically meaningful change is generally taken to be about ±0.30 log units (Pelli, Robson & Wilkins, 1988). A consumer-screen test, however carefully calibrated, is noisier than that.
A normal result does not rule out any condition, including those for which CSF is sometimes affected. Many of the things mold and chronic-illness patients describe (cognitive complaints, fatigue, mood changes, autonomic symptoms) are not what a contrast sensitivity test measures.
What the test can be useful for is change tracking on the same setup over time. If your environment changes — you move out of a water-damaged building, complete a remediation, or start a treatment protocol — and you retake on the same device with similar lighting in three months, the change in your curve is more informative than any single absolute number. That is the use the framework's own literature points to, and it is the use we would suggest, with all the caveats above.
Practical next steps
If you want a baseline:
- Take the test now. Three minutes for the quick mode, around seven for the full curve. Calibration happens at the start so the numbers are comparable across sessions on the same device.
- Save the result. Screenshot the curve or use the share-link generator. You want a record you can compare against later.
- Retake when something changes. Moved out of a building you suspect, finished a remediation, started or finished a treatment protocol, or simply want to check in three months — retake on the same device, similar lighting, similar viewing distance.
- Bring the trend to your clinicians. Whether your team is a CIRS-protocol practice, a neuro-optometrist, a functional-medicine clinician, or your primary care doctor — a curve over months adds context a single in-clinic measurement cannot reconstruct. Bring the spatial-frequency curve, not just a single number.
For further reading, the primer on what contrast sensitivity measures and Pelli-Robson vs FACT vs qCSF cover the methodology differences behind the FACT critique above.
References
- Pelli, D. G., Robson, J. G., & Wilkins, A. J. (1988). The design of a new letter chart for measuring contrast sensitivity. Clinical Vision Sciences, 2, 187–199. The Pelli-Robson chart paper. The ±0.15 log test-retest repeatability and ±0.30 log clinically-meaningful-change figures cited in the disclaimer derive from this measurement tradition.
- Pelli, D. G., & Bex, P. (2013). Measuring contrast sensitivity. Vision Research, 90, 10–14. Methodology review of the major contrast sensitivity instruments — ceiling effects, coarse step sizes, and reliability limitations of grating charts with discrete contrast steps (including the FACT family), plus the precision gains from adaptive procedures. The load-bearing citation for the methodology critique above.
- Lesmes, L. A., Lu, Z.-L., Baek, J., & Albright, T. D. (2010). Bayesian adaptive estimation of the contrast sensitivity function: the quick CSF method. Journal of Vision, 10(3):17. The qCSF method — a Bayesian adaptive procedure that estimates the entire CSF in a few minutes. The methodology bar for what a modern adaptive contrast sensitivity test looks like.
- Watson, A. B., & Pelli, D. G. (1983). QUEST: a Bayesian adaptive psychometric method. Perception & Psychophysics, 33(2), 113–120. The foundational efficient-threshold paper underlying modern adaptive contrast sensitivity testing.
- Mäntyjärvi, M., & Laitinen, T. (2001). Normal values for the Pelli-Robson contrast sensitivity test. Journal of Cataract and Refractive Surgery, 27(2), 261–266. Source for the age-stratified normative Pelli-Robson values that any home measurement needs to be read against.
A note on what we did not cite
The CIRS framework's primary literature is largely outside the conventional peer-reviewed channels these references sit in — patient-protocol books, clinic-network materials, conference proceedings. We have chosen not to selectively cite framework-internal papers without the methodology context they would require to evaluate fairly. The independent peer-reviewed literature on contrast sensitivity in mold-exposed or biotoxin-exposed cohorts is, as of the date of this post, smaller than the literature for any of the better-studied conditions we cover elsewhere. Where that gap closes, this post will be updated.