Study identifies that the sequence of previous diseases may help predict the risk of long COVID

Long COVID is a heterogeneous clinical condition that affects thousands of people and can manifest in many different ways. Understanding why some people develop it while others do not remains one of the main scientific challenges.

A new study led by the Germans Trias i Pujol Research Institute (IGTP) provides a new perspective: what matters is not only which previous diseases a person has, but also the order in which they appeared and how they interact. This approach makes it possible to identify risk profiles for long COVID that had not been detected until now. The research was conducted within the framework of the COVICAT study, coordinated in collaboration with the Barcelona Institute for Global Health (ISGlobal), a centre supported by the "la Caixa" Foundation.

The study, published in BMC Medicine, is based on data from more than 10,000 participants in the GCAT (Genomes for Life) cohort, which has collected clinical and genetic information from the Catalan population for over 15 years. Using these data, linked to the prospective COVID follow-up of the COVICAT study launched in 2020, the research team reconstructed health trajectories -that is, the temporal sequence of different chronic diseases- to analyse how these may influence the development of long COVID.

The importance of disease sequence

Until now, most studies had focused on whether or not a person had a previous condition. This work shows that the sequence and interaction of diseases over time can also be key to predicting the risk of developing long COVID.

"It is not enough to know which diseases a person has. The order in which they appear can significantly influence risk, especially among women," explains Natàlia Blay, first author of the study.

The results show that taking into account the sequence and interaction of diseases over time allows for a more accurate prediction than considering only the presence of a single condition. For example, individuals with anxiety followed by depression have a different risk compared with those who experience the same conditions in the reverse order. In total, 162 trajectories were analysed, and 38 were associated with a significantly higher risk of long COVID. The most frequent trajectories involved mental health disorders, neurological, respiratory (such as asthma), and metabolic or digestive diseases (such as hypertension, obesity or reflux).

The analysis also reveals that some of these disease trajectories increase the risk of long COVID regardless of the severity of the initial infection. This indicates that not everything can be explained by the type or intensity of acute COVID. The researchers note that, in the future, this approach could benefit from artificial intelligence tools capable of detecting complex patterns in large longitudinal health datasets, thus improving the ability to predict risks and identify vulnerable population groups more precisely.

"This work demonstrates that long COVID results from a prior health trajectory rather than a single factor. Above all, it highlights that studying trajectories in longitudinal data such as those from GCAT has value beyond COVID, as it allows us to identify population health patterns that may help predict other diseases and support a more preventive and personalised public health approach," explains Rafael de Cid, principal investigator of the study and director of GCAT at IGTP.

Regarding the genetic component, the study reveals no strong overall genetic correlation with long COVID, although modest relationships were found with genetic factors linked to neurological and musculoskeletal diseases. These findings suggest a possible shared susceptibility in certain cases. The study reinforces the need to understand health as a dynamic and cumulative process. Incorporating the temporal sequence of diseases alongside genetic information can improve prediction, care and prevention for long COVID and other chronic conditions.