Researchers reported 38 prepandemic disease trajectories that increased long COVID risk, potentially driven by mental health, neurologic, respiratory, and cardiometabolic-digestive patterns.

In the study, investigators assessed 8,322 adults aged 40 to 65 years in the Catalonia-based COVICAT cohort and identified long COVID in 12.4% (n = 1,029) of the participants. Long COVID prevalence reached 66.7% among adults with severe COVID-19 compared with 11.5% among those with nonsevere infections. Women showed higher long COVID rates (15.1% vs 8.5%), greater multimorbidity (mean 4.6 diagnoses vs 3.6), and all trajectory-level associations compared with men.

Using 15 years of electronic health records (2010 to 2019), the investigators reconstructed sequential patterns of 106 chronic conditions. Twenty-three conditions were associated with long COVID, and among 162 two-condition trajectories that were identified, 23 in the full cohort and 34 in women were associated with long COVID. Nearly half (45.6%) of these trajectories clustered within mental health–neuromuscular patterns, including depression, anxiety, migraine, headache syndromes, and stress. Respiratory and allergic pathways such as asthma and allergic rhinitis were also overrepresented. Cardiometabolic-digestive trajectories accounted for 19.3% of associations.

Temporality was clinically decisive: 73.7% of significant trajectories increased risk only when conditions occurred in a specific order. Symptom analysis showed that prior migraine corresponded to digestive complaints, severe stress to respiratory and fatigue symptoms, and mononeuropathies to neuropathic sensations. After adjustment for COVID-19–related hospitalization, 14 conditions and 15 trajectories remained independently associated with long COVID, whereas obesity and lipidaemia no longer met significance. A female-specific obesity followed by knee-derangement trajectory emerged. Among 2,108 genotyped participants, no statistically significant genetic correlations with long COVID were identified, although polygenic risk scores for intervertebral disc disorders and headache syndromes were associated with risk in women.

The study was limited by the relatively small number of long COVID cases, particularly among men, and by incomplete diagnosis prior to 2010 and private-sector health records. Use of broad ICD-10 groupings and evolving diagnostic criteria could have introduced misclassification, and mixed infection status among controls may have attenuated associations.

The authors reported no competing interests. Funding was provided by La Caixa Foundation, La Marató TV3, the Spanish Ministry of Health, the Spanish Ministry of Science and Innovation, and Horizon Europe END-VOC.

Source: BMC Medicine