Register

Cardiovascular diseases are the leading cause of death both in men and women worldwide. Heart failure (HF) is the most common form of heart disease, characterised by the heart’s inability to pump a sufficient supply of blood to meet the needs of the body. The lifetime risk of developing HF is roughly 20%, yet, it remains difficult to diagnose due to its and a lack of agreement of diagnostic criteria. As the diagnosis of HF is dependent on ascertainment of clinical histories and appropriate screening of symptomatic individuals, identifying those at risk of HF is essential. This DREAM challenge focuses on the prediction of HF using a combination of gut microbiome and clinical variables.

This challenge is designed to predict incident risk for heart failure in a large human population study of Finnish adults, FINRISK 2002 (Borodulin et al., 2018). The FINRISK study has been conducted in Finland to investigate the risk factors for cardiovascular disease every 5 years since 1972. A random sample stratified by sex and 10-year age-groups among the population aged 24–74 years from six geographical areas was taken. Of the 13,498 invitees, 8,783 participated in the study. Among the participants, 7,231 donated fecal samples (Salosensaari et al., 2021), which were then sequenced by shallow shotgun metagenomic sequencing for fecal microbiome characterization. National health care registers in Finland enable combining of the data in FINRISK with subsequent in- and outpatient disease diagnoses and drug prescriptions based on individual personal identity codes. This study protocol of FINRISK 2002 has been approved by the Coordinating Ethics Committee of the Helsinki University Hospital District (Helsinki, Finland) (ref. no. 558/E3/2001), and all participants provided written informed consent.

The aim of this challenge is time-to-event predictions of the heart failure using a combination of the information from fecal microbiome composition (via shallow shotgun metagenomic sequencing) and host phenotype including conventional risk factors, retrieved from population health registers. The metrics described in the “Assessment” section will be used to assess the performance in predicting heart failure.