Long-term trends in Post-COVID severity: a machine learning analysis from the POP/COVIDOM cohort of the German NAPKON Cohort Network - PubMed
7 hours ago
- #Machine learning
- #Post-COVID syndrome
- #Long COVID
- Post-COVID syndrome (PCS) affects survivors with varying symptom profiles, categorized by acute disease severity (PCS-S) or individual resilience (PCS-R).
- The study analyzed data from 1526 participants in the German NAPKON Cohort Network, tracking PCS-S and PCS-R scores over 9, 24, and 36 months.
- Findings show small but significant declines in PCS-S and PCS-R scores over time, indicating persistent symptom burden with gradual improvement.
- Predictive models explained 16.7-52.6% of variance in later PCS severity, with fatigue, age, quality of life, and depression as key predictors.
- Gender-specific predictors were identified: fatigue and sleep issues predicted PCS-R, with living/employment status relevant in females and cognitive deficits in males.
- The study highlights the need for tailored interventions for long-term management of PCS, given the stability of symptom severity over time.