Counterarguments and Epistemic Constraints
The chapter’s hidden-burden thesis — that ME/CFS is globally prevalent but systematically underdetected — has the structure of a claim that predicts its own invisibility. This section examines the structural limits of that claim and the evidence that would constrain or refute it.
The central argument — ME/CFS is prevalent but undetected in LMICs because diagnostic infrastructure is absent — has the logical structure of an unfalsifiable claim given current data infrastructure. Any study that fails to find ME/CFS in an LMIC setting can be attributed to: (a) the diagnostic criteria being culturally inappropriate, (b) the exclusionary laboratory workup being unavailable, (c) the study using the wrong triggering infection, or (d) the study excluding the most severely affected patients who cannot attend follow-up. The thesis has an unrestricted set of auxiliary hypotheses to absorb falsification.
This does not make the thesis wrong. It makes it non-falsifiable by methods short of population-based surveys with culturally adapted criteria, nutritional screening, orthostatic testing, and home-based severe-patient recruitment — a bar no HIC study has met either. But a non-falsifiable claim should carry lower certainty than its structural plausibility would otherwise warrant, because unfalsifiable claims are systematically overrepresented in scientific literatures (the “survivor bias” of hypotheses — falsifiable claims get tested and retired; unfalsifiable claims persist).
(Origin: brainstorm — Phase 4 categories 10 and 12.)
Consequence: “We can’t find it because we can’t look” is a reasonable statement. “We can’t find it, therefore it’s there” is not. The chapter’s claim should specify what evidence would change its mind — not as a rhetorical gesture, but as a methodological requirement. Severity applicability: all.
Three categories of evidence would substantially weaken or refute the chapter’s central claim:
Differential pathogen profiles. If post-infectious ME/CFS conversion rates vary by an order of magnitude across triggering infections — e.g., EBV → 5% CFS, dengue → 0.5%, TB → ~0.1% — then extrapolating from HIC herpesvirus cohorts to LMIC arbovirus/TB cohorts is quantitatively wrong. Chikungunya data (38% chronic fatigue at 30 months) argue against large variation, but no head-to-head comparison of pathogen-specific ME/CFS conversion rates exists.
Protective factors in high-infectious-disease settings. Chronic helminth co-colonization induces Th2/Treg immune dominance, which suppresses the Th1/Th17 responses implicated in post-infectious autoimmunity. Early-life exposure to high infectious-disease burden may train the immune system away from hyperinflammatory post-infectious responses (“old friends” hypothesis). Younger population age structure in LMICs mechanically reduces age-adjusted ME/CFS prevalence. If any or all of these protective mechanisms operate, ME/CFS prevalence may be genuinely lower in LMICs — not because diagnostic access is worse, but because susceptibility is lower.
Competing mortality. Patients who would develop post-infectious ME/CFS in an HIC, where ICU care is available, die of the acute infection in an LMIC where it is not — removing susceptible individuals from the denominator before they can develop the chronic condition.
None of these counterarguments has strong empirical support. They are evidential mirror images of the hidden-burden thesis: both rest on inference from the same absence of data, in opposite directions.
(Origin: brainstorm — Phase 4 categories 10 and 11. Certainty: 0.20 — counterarguments are plausible but empirically unsupported.)
Consequence: It is possible — not likely, but possible — that ME/CFS is genuinely rarer in LMICs. The paper should be transparent that both “hidden burden” and “genuinely lower prevalence” are hypotheses on equal evidential footing until sentinel surveillance data exist. Severity applicability: all.
ME/CFS is coded as G93.3 in ICD-10 and 8E49 in ICD-11 — the WHO’s classification expansion is a global-level policy achievement. But the pathway from ICD code to patient benefit requires: national health information systems adopting ICD-11, physicians trained to recognize ME/CFS and assign the code, diagnostic criteria deployable without laboratory exclusion workup, disability systems that recognize the code and provide benefits, and clinical management guidelines that give physicians something to do after assigning the code. In most LMICs, none of these conditions hold. The code exists on paper. It changes nothing on the ground.
(Origin: brainstorm. Certainty: 0.40 — implementation-gap argument is structural, drawing on well-documented patterns in global-health policy. No study has audited ICD-11 8E49 adoption rates by country. The “empty vessel” framing is a hypothesis, not an established finding.)
Consequence: Getting ME/CFS into ICD-11 was a victory. But if no one in the countries with the most patients can use the code, the victory is a monument, not a functioning door. Severity applicability: all — the implementation gap affects all severity levels, though the severely disabled have the most to lose from a non-functional diagnostic label.
Falsifiable prediction: If a systematic audit of ≥10 LMIC national health systems finds ≥3 where ICD-11 8E49 has been adopted into the national health information system AND appears in ≥1% of primary care encounter records, the empty-vessel hypothesis is refuted. If zero countries meet these conditions 5 years post-ICD-11 adoption, the hypothesis is supported.