Evidence is weighed, not counted
Weak evidence does not become strong because it points the other way. A tiny case report, a broad environmental-clinic referral sample, or a conditioning experiment in healthy volunteers cannot be given the same causal weight as a well-defined MCS cohort, a randomized trial, an occupational exposure study, or a reproducible physiological finding. The question is not whether a paper sounds contrary to our position. The question is what its design actually allows it to prove.
The 2006 Das-Munshi review can support a modulation claim; it cannot establish psychological origin. It reviewed 37 provocation studies and found that positive reactions were common when active and sham conditions could be distinguished, while masked or below-odor-threshold studies were more often negative. That is good evidence that expectation and sensory recognition influence symptom expression after illness exists. It does not show what caused the illness to begin.
Conditioning experiments show that human nervous systems can learn symptom associations; they do not show that learned association initiated MCS. Leuven researchers paired odors with CO2 inhalation in healthy participants and showed that an odor alone could later evoke symptoms, that learned symptoms could generalize to new odors, and that expectancy influenced the effect. These are elegant demonstrations of learning after a real physiological challenge—not demonstrations of MCS onset.
But most of those conditioning experiments were not experiments in people whose MCS onset was being observed. They demonstrate a mechanism the human nervous system possesses. They do not show that this mechanism initiated consensus-defined MCS in the community.
Psychiatric comorbidity cannot establish psychiatric causation. Severe chronic MCS can produce isolation, exposure vigilance, depression, anxiety, and loss of safety. Some referral-clinic studies reported high psychiatric rates, but several used broad environmental-illness populations rather than clean consensus-defined MCS cohorts. Studying who arrives at a psychosomatic or environmental clinic is not the same as establishing what causes MCS.
The controlled treatment evidence also fails the stronger psychological-causation prediction. If maladaptive beliefs or cognitive-emotional representations were the central maintaining cause, changing them should substantially improve MCS itself. In the strongest MCS-specific randomized trial, mindfulness-based cognitive therapy changed illness perceptions but did not improve overall MCS illness status. That does not make psychological support useless; it directly weakens the claim that changing thought patterns reverses the disorder.
Put together, the psychological literature explains part of the response architecture: expectation, conditioning, fear, attention, and learned association can modify symptom expression. It does not adequately explain environmental onset patterns, measurable sensory physiology, autonomic and inflammatory findings, or the failure of cognitive change to reliably restore chemical tolerance.