The Evidence

What the 2006 provocation review actually found

The provocation literature shows that recognition, expectation, masking, and prior beliefs can change symptom responses. It does not establish that psychological processes initiated consensus-defined MCS. The 2006 review is evidence for modulation and a warning about poor challenge-study design—not proof that MCS is psychogenic.

The provocation literature is often cited as though it settled MCS as psychological. It did not. Its strongest finding is that recognition and expectation can modify reactions during laboratory challenges after chemical sensitivity already exists.

What we know

A provocation study sounds simple: expose a chemically sensitive person to a suspected trigger and to a sham condition, blind the participant, and see whether symptoms follow the real exposure. In principle, this is exactly the kind of experiment MCS needs. In practice, it is exceptionally difficult to blind an exposure that has an odor, irritant quality, or other detectable cue.

The 2006 Das-Munshi review is the paper most often used to support a psychogenic interpretation, but its design question was narrower: can symptoms be reproduced during challenge when active and sham exposures are masked? It did not observe the onset of MCS, and blinding was inadequate in most of the studies it reviewed.

What the evidence shows

The review’s results matter, but they must be interpreted inside those limits. It identified 37 provocation studies involving 784 people reporting MCS, 547 controls, and 180 additional participants of whom a subset were chemically sensitive. Blinding was inadequate in most studies.

When the active exposure could probably be recognized, positive responses were common; when recognition was reduced, chemical-specific responses became much harder to demonstrate. In 21 studies the active chemical was probably detectable by odor and 19 reported positive responses among chemically sensitive participants. In seven studies using exposures at or below odor thresholds, six failed to show consistent responses. Six forced-choice studies found that chemically sensitive participants were not better than controls at detecting odor thresholds, and three studies using masking agents found no association between active provocation and response.

What those results support is modulation: recognition, expectation, perception, and learned association can influence symptom expression during a laboratory challenge. They do not show what initiated MCS, whether illness began before environmental attribution, or whether changing beliefs restores chemical tolerance.

But the review does not establish several stronger claims often built on top of it. It does not show what initiated MCS. It does not show that a participant was healthy before developing beliefs about chemicals. It does not show that thought patterns can reverse consensus-defined MCS. And it does not establish that every biologically relevant trigger should produce an immediate measurable response during a brief chamber challenge.

MCS symptoms can be delayed, multisystem, and individualized. Real-world exposures can be mixtures rather than single chemicals. A challenge dose high enough to be biologically meaningful may also be detectable, breaking the blind. A dose low enough to preserve blinding may fail to engage the relevant pathway. These are not excuses for positive studies; they are design problems that future studies have to solve.

A later critique by Goudsmit and Howes argued that the 2006 review overinterpreted parts of the literature and that participant selection and methodology weakened the case for a learned-response explanation. We do not need that critique to dismiss the review. We do not dismiss it. We assign it the conclusion it actually supports.

What the registry is answering

The registry provides something chamber studies usually lack: chronology and individualized trigger history. It can identify participants with repeated, well-characterized reactions, including episodes occurring before the person knew a trigger was present, and recruit those individuals into better blinded studies.

It can also capture latency and duration. That allows future provocation work to measure outcomes over the time window in which each participant actually reports reactions rather than assuming all meaningful physiology occurs in the first few minutes.

Where we stand

The provocation literature shows that expectation and stimulus recognition can shape MCS reactions. That is important evidence for modulation and conditioning.

It does not establish psychological initiation of MCS. The correct conclusion is narrower and stronger: a serious MCS model must include both bottom-up biological triggering and top-down modulation, and the next generation of provocation studies has to be designed well enough to separate them.

We followed the evidence. Here’s where it leads →

Sources

Sources

1Das-Munshi J, Rubin GJ, Wessely S. Multiple chemical sensitivities: a systematic review of provocation studies. Journal of Allergy and Clinical Immunology. 2006;118(6):1257–1264. PMID: 17137865.
2Goudsmit E, Howes S. J Nutr Environ Med 2008.
3Goudsmit E, Howes S. Is multiple chemical sensitivity a learned response? A critical evaluation of provocation studies. 2008.
4Van den Bergh O, Stegen K, Van Diest I, et al. Acquisition and extinction of somatic symptoms in response to odours: a Pavlovian paradigm relevant to multiple chemical sensitivity. Occupational and Environmental Medicine. 1999;56(5):295–301. PMID: 10472302.
← Back to MCSDisability.com
↑