← PAIS Cohort Database

Susceptibility & conversion predictors

What predicts who develops the chronic syndrome given infection (conversion), with emphasis on variables measured during the acute phase. Four causal questions — susceptibility to infection, to severe acute disease, to conversion, and to non-recovery — are kept strictly separate and never aggregated. A predictor measured after the outcome began is not counted as temporality-valid, which excludes most of what the literature calls a "risk factor". No effect is pooled across pathogens: different contrasts and adjustment sets are not comparable numbers.

17 predictors · 5 null results (29%) · 14 temporality-valid · harmonisation ruleset pais-severity-harmony-v1 · factor registry · predictors CSV · spec · CC BY 4.0

Null results are recorded, not dropped — they are 29% of this table and the reason to build this rather than read a review. Factors are ranked by replication and temporality validity, never by effect size.

Replication table

One row per (factor or harmonised construct × outcome × question), sorted by how many cohorts tested it. The two rows that stand out — acute severity (replicated) and the long tail tested exactly once — are the finding.

One row per (factor/harmonised-construct × outcome × question). Sorted by replication (cohorts tested), showing null results alongside significant ones. No pooled effect is computed — different contrasts and adjustment sets are not poolable.

Factor / constructOutcomeQuestionCohortsPathogensSignificantNullSign concordanceTemporality-validMedian predictors tested
acute:illness-severity-compositeAny persistent symptomConversion (given infection → syndrome)11101/11/1n/r
immune:convalescent-antibody-titreAny persistent symptomConversion (given infection → syndrome)11101/10/1n/r
prior:chronic-lung-diseaseAny persistent symptomConversion (given infection → syndrome)11101/11/1n/r
host:sex-femaleFatigueConversion (given infection → syndrome)11101/10/1n/r
gene:rabgap1lChronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11101/11/1n/r
gene:fbxl4Chronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11101/11/1n/r
gene:olfm4Chronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11101/11/1n/r
gene:ca10Chronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11101/11/1n/r
acute:illness-severity-compositeChronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11101/11/1n/r
gene:foxp4-locusLong COVID / post-COVID condition (caseness)Conversion (given infection → syndrome)11101/11/1n/r
gene:hla-dqb1-0602Narcolepsy (incident)Conversion (given infection → syndrome)11101/11/11
host:ageArthralgia / rheumatic musculoskeletal painConversion (given infection → syndrome)11101/10/1n/r
host:ageChronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11011/1n/r
host:sex-femaleChronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11011/1n/r
prior:premorbid-psychiatricChronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11011/1n/r
pathogen:infecting-agentChronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11011/1n/r
immune:acute-cytokinesChronic fatigue syndrome / ME-CFS casenessConversion (given infection → syndrome)11011/1n/r

Predictors by disease

A separate subsection for each disease/trigger — everything tested in that syndrome's cohorts, significant and null, with contrast, measurement window and temporality per row.

Jump to disease: Chikungunya virus · Dengue virus · SARS-CoV-2 · Pandemrix (AS03-adjuvanted H1N1 influenza vaccine) · Mixed (EBV / Ross River virus / Coxiella burnetii) · Unknown / idiopathic trigger

Chikungunya virus Virus 1 predictor

1 significant · 0 null · 0 temporality-valid · from 1 cohort

FactorContrastWindowOutcomeDirectionQuestionModelTemporalityCohort
Ageolder age (per decade / older vs younger)Pre-infectionArthralgia / rheumatic musculoskeletal pain @ 24mo↑ increases riskConversion (given infection → syndrome)multivariable logisticinvalidTELECHIK cohort

Dengue virus Virus 1 predictor

1 significant · 0 null · 0 temporality-valid · from 1 cohort

FactorContrastWindowOutcomeDirectionQuestionModelTemporalityCohort
Female sexfemale vs male (among dengue patients)Genetic / fixedFatigue @ 2mo↑ increases riskConversion (given infection → syndrome)univariableinvalidColombo Dengue Study

SARS-CoV-2 Virus 4 predictors

4 significant · 0 null · 3 temporality-valid · from 2 cohorts

FactorContrastWindowOutcomeDirectionQuestionModelTemporalityCohort
Acute illness severity (composite index) → Acute severityhigher acute illness severityAcute phaseAny persistent symptom @ 6mo↑ increases riskConversion (given infection → syndrome)multivariable logisticvalidBergen COVID-19 home-isolated cohort
Convalescent antibody titrehigher convalescent anti-SARS-CoV-2 antibody titreEarly convalescent (≤3mo)Any persistent symptom @ 6mo↑ increases riskConversion (given infection → syndrome)multivariable logisticinvalidBergen COVID-19 home-isolated cohort
Pre-existing chronic lung diseasepre-existing chronic lung disease vs nonePre-infectionAny persistent symptom @ 6mo↑ increases riskConversion (given infection → syndrome)multivariable logisticvalidBergen COVID-19 home-isolated cohort
FOXP4 locus (rs9367106)rs9367106 risk allele vs reference (per-allele)Genetic / fixedLong COVID / post-COVID condition (caseness)↑ increases riskConversion (given infection → syndrome)mixed effectsvalidLong COVID Host Genetics Initiative GWAS

Pandemrix (AS03-adjuvanted H1N1 influenza vaccine) Vaccine 1 predictor

1 significant · 0 null · 1 temporality-valid · from 1 cohort

FactorContrastWindowOutcomeDirectionQuestionModelTemporalityCohort
HLA-DQB1*06:02 carriageHLA-DQB1*06:02 positive vs negativeGenetic / fixedNarcolepsy (incident)↑ increases riskConversion (given infection → syndrome)NonevalidFinland Pandemrix childhood narcolepsy cohort

Mixed (EBV / Ross River virus / Coxiella burnetii) Mixed 6 predictors

1 significant · 5 null · 6 temporality-valid · from 1 cohort

FactorContrastWindowOutcomeDirectionQuestionModelTemporalityCohort
Acute illness severity (composite index) → Acute severityhigher acute illness severity (per unit of the acute symptom-severity score)Acute phaseChronic fatigue syndrome / ME-CFS caseness @ 6mo↑ increases riskConversion (given infection → syndrome)multivariable logisticvalidDubbo Infection Outcomes Study
Ageper year of agePre-infectionChronic fatigue syndrome / ME-CFS caseness @ 6monull resultConversion (given infection → syndrome)multivariable logisticvalidDubbo Infection Outcomes Study
Female sexfemale vs maleGenetic / fixedChronic fatigue syndrome / ME-CFS caseness @ 6monull resultConversion (given infection → syndrome)multivariable logisticvalidDubbo Infection Outcomes Study
Premorbid psychiatric historypremorbid psychiatric history vs nonePre-infectionChronic fatigue syndrome / ME-CFS caseness @ 6monull resultConversion (given infection → syndrome)multivariable logisticvalidDubbo Infection Outcomes Study
Infecting pathogen (which agent)EBV vs Ross River virus vs Coxiella burnetiiAcute phaseChronic fatigue syndrome / ME-CFS caseness @ 6monull resultConversion (given infection → syndrome)univariablevalidDubbo Infection Outcomes Study
Acute cytokine levels (IL-6/IL-1β/IFN-γ/TNF-α/IL-10) → Inflammatory burdenhigher acute cytokine levelsAcute phaseChronic fatigue syndrome / ME-CFS caseness @ 6monull resultConversion (given infection → syndrome)univariablevalidDubbo Infection Outcomes Study

Unknown / idiopathic trigger Unknown 4 predictors

4 significant · 0 null · 4 temporality-valid · from 1 cohort

FactorContrastWindowOutcomeDirectionQuestionModelTemporalityCohort
RABGAP1L locusrisk allele vs reference at the RABGAP1L signalGenetic / fixedChronic fatigue syndrome / ME-CFS caseness↑ increases riskConversion (given infection → syndrome)mixed effectsvalidDecodeME ME/CFS GWAS
FBXL4 locusrisk allele vs reference at the FBXL4 signalGenetic / fixedChronic fatigue syndrome / ME-CFS caseness↑ increases riskConversion (given infection → syndrome)mixed effectsvalidDecodeME ME/CFS GWAS
OLFM4 locusrisk allele vs reference at the OLFM4 signalGenetic / fixedChronic fatigue syndrome / ME-CFS caseness↑ increases riskConversion (given infection → syndrome)mixed effectsvalidDecodeME ME/CFS GWAS
CA10 locusrisk allele vs reference at the CA10 signalGenetic / fixedChronic fatigue syndrome / ME-CFS caseness↑ increases riskConversion (given infection → syndrome)mixed effectsvalidDecodeME ME/CFS GWAS

Factor × pathogen matrix

Temporality-valid predictors only (pre-infection / acute-phase / genetic, in prospective-inception or registry cohorts). Only domains with at least one tested factor are shown; the empty cells within them are the point — most acute-laboratory factors are never tested against the chronic outcome.

tested, significant tested, null tested, conflicting · never tested
FactorSARS-CoV-2Borrelia burgdorferi (Lyme disease)Giardia duodenalis (lamblia)Zika virusDengue virusUnknown / idiopathic triggerMixed (EBV / Ross River virus / Coxiella burnetii)Epstein-Barr virus (EBV)SARS-CoV-1COVID-19 vaccine (mRNA / spike-based)Chikungunya virusCoxiella burnetii (Q fever)Ross River virusPandemrix (AS03-adjuvanted H1N1 influenza vaccine)Ebola virusWest Nile virusCampylobacter jejuni / E. coli O157:H7 (Walkerton outbreak)
Host demographic
Age······null··········
Female sex······null··········
Body mass index / obesity·················
Host genetic
HLA type·················
Cytokine gene polymorphism·················
GWAS susceptibility locus (generic)·················
HLA-DQB1*06:02 carriage·············sig···
FOXP4 locus (rs9367106)sig················
RABGAP1L locus·····sig···········
OLFM4 locus·····sig···········
FBXL4 locus·····sig···········
CA10 locus·····sig···········
Host prior health
Pre-existing comorbidity count·················
Pre-existing chronic lung diseasesig················
Prior ME/CFS or fibromyalgia·················
Atopy / allergy·················
Premorbid psychiatric history······null··········
Acute clinical
Acute illness severity (composite index)sig·····sig··········
Hospitalisation for acute illness·················
ICU admission·················
Fever duration (days)·················
Peak temperature·················
Symptom count at presentation·················
Supplemental oxygen requirement·················
Acute neurological involvement·················
WHO warning signs (dengue)·················
Time from onset to presentation·················
Acute pathogen
Viral load / Ct value·················
NS1 antigenaemia (dengue)·················
Dengue serotype·················
Primary vs secondary infection (serostatus)·················
Variant / strain·················
Duration of viraemia / shedding·················
Antigen persistence at convalescence·················
Infecting pathogen (which agent)······null··········
Acute immune
Acute cytokine levels (IL-6/IL-1β/IFN-γ/TNF-α/IL-10)······null··········
Convalescent antibody titre·················
Autoantibody presence·················

Acute-phase coverage

Which cohorts analysed any acute-phase factor against the chronic outcome (✓), collected but did not analyse it (c), or neither (·). The "c" cells are recoverable through collaboration.

CohortAcute clinicalAcute laboratoryAcute pathogenAcute treatmentAcute immune
Bergen COVID-19 home-isolated cohort···
Dubbo Infection Outcomes Study··

Every estimate, by factor

All estimates for each factor across cohorts, with contrast, measurement window, adjustment set and temporality shown per row.

Acute illness severity (composite index) Acute clinical → Acute severity

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Bergen COVID-19 home-isolated cohortSARS-CoV-2higher acute illness severityAcute phase↑ increases riskmultivariable logisticage, sexvalid
Dubbo Infection Outcomes StudyMixed (EBV / Ross River virus / Coxiella burnetii)higher acute illness severity (per unit of the acute symptom-severity score)Acute phase↑ increases riskmultivariable logisticage, sex, psychological factors, infecting agentvalid

Infecting pathogen (which agent) Acute pathogen

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Dubbo Infection Outcomes StudyMixed (EBV / Ross River virus / Coxiella burnetii)EBV vs Ross River virus vs Coxiella burnetiiAcute phasenull resultunivariablevalid

Acute cytokine levels (IL-6/IL-1β/IFN-γ/TNF-α/IL-10) Acute immune → Inflammatory burden

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Dubbo Infection Outcomes StudyMixed (EBV / Ross River virus / Coxiella burnetii)higher acute cytokine levelsAcute phasenull resultunivariablevalid

Convalescent antibody titre Acute immune

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Bergen COVID-19 home-isolated cohortSARS-CoV-2higher convalescent anti-SARS-CoV-2 antibody titreEarly convalescent (≤3mo)↑ increases riskmultivariable logisticage, sexinvalid

Age Host demographic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Dubbo Infection Outcomes StudyMixed (EBV / Ross River virus / Coxiella burnetii)per year of agePre-infectionnull resultmultivariable logisticvalid
TELECHIK cohortChikungunya virusolder age (per decade / older vs younger)Pre-infection↑ increases riskmultivariable logisticsexinvalid

Female sex Host demographic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Colombo Dengue StudyDengue virusfemale vs male (among dengue patients)Genetic / fixed↑ increases riskunivariableinvalid
Dubbo Infection Outcomes StudyMixed (EBV / Ross River virus / Coxiella burnetii)female vs maleGenetic / fixednull resultmultivariable logisticvalid

Pre-existing chronic lung disease Host prior health

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Bergen COVID-19 home-isolated cohortSARS-CoV-2pre-existing chronic lung disease vs nonePre-infection↑ increases riskmultivariable logisticage, sexvalid

Premorbid psychiatric history Host prior health

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Dubbo Infection Outcomes StudyMixed (EBV / Ross River virus / Coxiella burnetii)premorbid psychiatric history vs nonePre-infectionnull resultmultivariable logisticvalid

HLA-DQB1*06:02 carriage Host genetic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Finland Pandemrix childhood narcolepsy cohortPandemrix (AS03-adjuvanted H1N1 influenza vaccine)HLA-DQB1*06:02 positive vs negativeGenetic / fixed↑ increases riskNonevalid

FOXP4 locus (rs9367106) Host genetic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
Long COVID Host Genetics Initiative GWASSARS-CoV-2rs9367106 risk allele vs reference (per-allele)Genetic / fixed↑ increases riskmixed effectsgenetic principal components, severe acute COVID-19 signalvalid

RABGAP1L locus Host genetic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
DecodeME ME/CFS GWASUnknown / idiopathic triggerrisk allele vs reference at the RABGAP1L signalGenetic / fixed↑ increases riskmixed effectsgenetic principal componentsvalid

OLFM4 locus Host genetic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
DecodeME ME/CFS GWASUnknown / idiopathic triggerrisk allele vs reference at the OLFM4 signalGenetic / fixed↑ increases riskmixed effectsgenetic principal componentsvalid

FBXL4 locus Host genetic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
DecodeME ME/CFS GWASUnknown / idiopathic triggerrisk allele vs reference at the FBXL4 signalGenetic / fixed↑ increases riskmixed effectsgenetic principal componentsvalid

CA10 locus Host genetic

CohortPathogenContrastWindowDirectionModelAdjusted forTemporality
DecodeME ME/CFS GWASUnknown / idiopathic triggerrisk allele vs reference at the CA10 signalGenetic / fixed↑ increases riskmixed effectsgenetic principal componentsvalid