{
  "abstract": "Background In England, one-third of adults have diagnosed high blood pressure, and about two-thirds of them develop additional comorbidities. This significantly reduces quality of life. With an aging population and rising prevalence of hypertension, it's crucial to identify likely comorbidities to develop in individuals with high blood pressure. Investigating comorbidity patterns across different hypertension subgroups (e.g., diagnosed vs. undiagnosed, treated vs. untreated) is also of interest, as there in not much research in this area.This understanding enables clinicians to predict, manage, and prevent such comorbidities. This study aims to identify comorbidity risks in hypertensive individuals and explore variations across subgroups to inform targeted preventions and management strategies.Methods We initially identified hypertensive participants from the UK Biobank (UKB), defining whether they were previously diagnosed, on treatment, and had achieved controlled BP using baseline measurements, self-reported information, and health records. Using a series of ICD-9/10 codes for automatic phenotyping of disease status (phecodes), we identified the occurrence of incident comorbidities. These data were used perform a phenome-wide association study (PheWAS) to establish the incidence and relative risk of comorbidity development within each subgroup. Additionally, we determined the censored incidence rate and relative risk using cox regression and Kaplan-Meier curves. All of this was visualised using PheWAS graphs, bar charts, and box plots.Findings We identified 1116 statistically significant phenotypes within the any hypertension sub cohort, 1180 within the diagnosed hypertension sub cohort, 950 within the treated sub cohort, and only 14 within the controlled hypertension sub cohort (refer to figure 1) . Our PheWAS analysis showed that all known comorbidities have a HR of greater than 1, with none of the 95% CIs crossing the dashed line (refer to figure 2). This suggests all these known phenotypes have a higher risk of developing in individuals with any type of hypertension compared to someone who may have normotensive BP readings. Figure 3 illustrates that a significant proportion of phenotypes (67%) exceeded the significance threshold for diagnosed and undiagnosed hypertension. The top 20 significantly associated phenotypes (lowest p-value) are clustered in the following body systems: endocrine/metabolic (3 phenotypes), circulatory system (10 phenotypes), Musculoskeletal (1 phenotype), genitourinary (2 phenotypes), digestive (1 phenotype) and respiratory (3 phenotype).Abstract 7-018 Figure 1Abstract 7-018 Figure 2Abstract 7-018 Figure 3Our PheWAS analyses revealed significant associations that have not been previously mentioned in existing literature. These encompassed various diseases and disease systems, such as phenotypes observed with neurological diseases, neoplasms, and the musculoskeletal system.Interpretation Understanding the causality behind the observed associations through Mendelian randomisation (MR) studies is crucial. It helps determine whether the identified comorbidity patterns are a direct result of hypertension or coincidental. This clarification guides medical decisions, enables targeted treatments, enhances study credibility, and contributes to a broader understanding of physiological interactions.",
  "authors": [
    {
      "affiliations": [
        "The Victor Phillip Dahdaleh Heart & Lung Research Institute, University of Cambridge, Cambridge, UK"
      ],
      "name": "Bisam-Ul Haq"
    },
    {
      "affiliations": [
        "The Victor Phillip Dahdaleh Heart & Lung Research Institute, University of Cambridge, Cambridge, UK"
      ],
      "name": "Samuel Lambert"
    },
    {
      "affiliations": [
        "Acute General Medicine, John Radcliffe Hospital, Oxford University Hospitals NHS Foundation Trust, UK"
      ],
      "name": "Mustafa Javaid"
    }
  ],
  "title": "7-018 Predicting multimorbidity risk in hypertensive individuals: understanding disease incidence",
  "uid": "27ab084c-1efc-5fea-ab1f-c141efc9c771"
}