Deep learning analysis of long COVID and vaccine impact in low- and middle-income countries (LMICs): development of a risk calculator in a multicentric study. — CFSMEATLAS
E2 ModerateModerate confidencePEM not requiredCross-SectionalPeer-reviewedMachine draft
Standard · 3 min
Deep learning analysis of long COVID and vaccine impact in low- and middle-income countries (LMICs): development of a risk calculator in a multicentric study.
Shaheen, Ahmed, Shaheen, Nour, Long COVID Collaboration Study Group in the LMICs et al. · Frontiers in public health · 2025 · DOI
Quick Summary
Researchers studied over 2,400 COVID-19 patients in low- and middle-income countries to understand why some people develop long-lasting symptoms like chronic fatigue and depression after infection. They used advanced computer analysis to identify which patients were at highest risk for these prolonged symptoms. The study found that vaccination reduced the risk of developing these long-term problems, and that older age, being female, and smoking were linked to higher risk of chronic fatigue.
Why It Matters
This study addresses a critical gap in understanding long COVID and CFS in underrepresented populations (LMICs), where research is scarce. The development of risk calculators could help clinicians identify vulnerable patients early for targeted intervention, and the confirmation of vaccination's protective effect provides important epidemiological evidence for post-pandemic public health strategies.
Observed Findings
68.1% of patients experienced symptoms lasting longer than 2 weeks post-COVID-19 infection
6.5% of the cohort met chronic fatigue syndrome (CFS) criteria
47.7% of patients reported depression symptoms
23.7% of vaccinated individuals reported infection after vaccination
Vaccinated individuals showed significantly lower odds of prolonged COVID-19 symptoms, CFS, and depression compared to unvaccinated individuals
Inferred Conclusions
Vaccination provides protective benefits against prolonged COVID-19 symptoms, CFS, and depression post-infection
Sex, age, and smoking status are associated risk factors for CFS development after COVID-19
Deep learning and machine learning approaches can successfully identify high-risk individuals for chronic post-COVID complications
The burden of long-term sequelae is substantial in LMICs and warrants comprehensive management strategies
Remaining Questions
Why is the observed CFS prevalence (6.5%) substantially lower than reported in other COVID-19 cohorts, and does this reflect true epidemiology or diagnostic methodology differences?
What This Study Does Not Prove
The cross-sectional design cannot prove that vaccination prevents long COVID/CFS development—it only shows an association. The study cannot determine whether protective factors cause reduced risk or whether reverse causality or selection bias explains the associations. Additionally, the low CFS prevalence may reflect misclassification rather than true disease burden in these populations.
Tags
Symptom:Cognitive DysfunctionFatigue
Phenotype:Infection-TriggeredLong COVID Overlap
Method Flag:PEM Not DefinedWeak Case DefinitionMixed Cohort
About the PEM badge: “PEM required” means post-exertional malaise was an explicit required diagnostic criterion for participant inclusion in this study — not that PEM was studied, observed, or discussed. Studies using criteria that do not require PEM (e.g. Fukuda, Oxford) are tagged “PEM not required”. How the atlas works →
What are the specific mechanisms by which vaccination reduces risk of prolonged symptoms and CFS?
How do sociocultural and healthcare access differences across LMICs influence symptom reporting and diagnosis rates?
Do the identified risk factors (sex, age, smoking) interact with vaccination status, and what is their predictive value for individual patient counseling?