Socio-Economic Risk Factors for Long COVID the Predi-COVID Cohort of Luxembourg

Background: Long COVID, defined as unexplained persistent symptoms experienced after COVID-19 infection, is a complex multi-organ condition with unclear aetiology but considerable societal implications. While research on pathophysiological mechanisms underlying established medical risk factors is ongoing, a literature on socio-economic determinants has emerged but remains limited in geographical scope and quality. We enrich the evidence by reporting an interpretable battery of patient-level socio-economic, demographic and medical risk factors for 12-month Long COVID, in Luxembourg’s Predi-COVID cohort study.

Methods: Participants, enrolled after PCR-confirmed SARS-CoV-2 infection, completed comprehensive questionnaires at baseline and 12-month follow-up covering socio-demographics, medical history, and persistent symptoms. As a primary outcome, Long COVID was defined as occurrence of any persistent symptom at follow-up, while a secondary outcome based on at least five symptoms was defined to explore robustness to symptom misclassification. Candidate risk factors included age, sex, low-income status, family size, educational attainment, employment, comorbidities, body mass index and, for a sub-sample, vaccination status. Risk ratios were estimated using quasi-Poisson regression, reweighted to Luxembourg’s age-sex structure.

Findings: The main analytical sample, made of 585 complete cases (54·4%), is younger and slightly skewed towards males vis-à-vis the general population, with 237 (40·51%) and 135 (23·07%) patients reporting at least one or five persisting symptoms, respectively. Analyses replicate established findings on the role of sex, age, BMI, comorbidities, and vaccination. They also identify a protective role of higher educational attainment: conditional on other factors, post-secondary and tertiary education are associated with a 30% (IRR: 0·70; 95%CI: 0·50–0·97) and 20% (IRR: 0·79; 95%CI: 0·64–0·97) lower risk of Long COVID. Associations were robust to the use of the secondary outcome.

Interpretation: In this test-positive, re-weighted cohort, higher educational attainment was independently protective against Long COVID at 12-month, alongside established demographic and medical risk factors. The findings support a role for social determinants, suggesting low education as a possible marker of life-course chronic inflammation and antecedent of inflammatory processes in Long COVID.