Dawit T Zemedikun1, Laura J Gray2, Kamlesh Khunti3, Melanie J Davies3, Nafeesa N Dhalwani4. 1. Health Economics Unit, Institute of Applied Health Research, University of Birmingham, Edgbaston, Birmingham, UK; Department of Health Sciences, Centre for Medicine, University of Leicester, Leicester, UK. 2. Department of Health Sciences, Centre for Medicine, University of Leicester, Leicester, UK. 3. Diabetes Research Centre, University of Leicester, Leicester General Hospital, Leicester, UK. 4. Diabetes Research Centre, University of Leicester, Leicester General Hospital, Leicester, UK. Electronic address: nnd2@le.ac.uk.
Abstract
OBJECTIVE: To assess the prevalence, disease clusters, and patterns of multimorbidity using a novel 2-stage approach in middle-aged and older adults from the United Kingdom. PATIENTS AND METHODS: Data on 36 chronic conditions from 502,643 participants aged 40 to 69 years with baseline measurements between March 13, 2006, and October 1, 2010, from the UK Biobank were extracted. We combined cluster analysis and association rule mining to assess patterns of multimorbidity overall and by age, sex, and ethnicity. A maximum of 3 clusters and 30 disease patterns were mined. Comparisons were made using lift as the main measure of association. RESULTS: Ninety-five thousand seven hundred-ten participants (19%) had 2 or more chronic conditions. The first cluster included only myocardial infarction and angina (lift=13.3), indicating that the likelihood of co-occurrence of these conditions is 13 times higher than in isolation. The second cluster consisted of 26 conditions, including cardiovascular, musculoskeletal, respiratory, and neurodegenerative diseases. The strongest association was found between heart failure and atrial fibrillation (lift=23.6). Diabetes was at the center of this cluster with strong associations with heart failure, chronic kidney disease, liver failure, and stroke (lift>2). The third cluster contained 8 highly prevalent conditions, including cancer, hypertension, asthma, and depression, and the strongest association was observed between anxiety and depression (lift=5.0). CONCLUSION: Conditions such as diabetes, hypertension, and asthma are the epicenter of disease clusters for multimorbidity. A more integrative multidisciplinary approach focusing on better management and prevention of these conditions may help prevent other conditions in the clusters.
OBJECTIVE: To assess the prevalence, disease clusters, and patterns of multimorbidity using a novel 2-stage approach in middle-aged and older adults from the United Kingdom. PATIENTS AND METHODS: Data on 36 chronic conditions from 502,643 participants aged 40 to 69 years with baseline measurements between March 13, 2006, and October 1, 2010, from the UK Biobank were extracted. We combined cluster analysis and association rule mining to assess patterns of multimorbidity overall and by age, sex, and ethnicity. A maximum of 3 clusters and 30 disease patterns were mined. Comparisons were made using lift as the main measure of association. RESULTS: Ninety-five thousand seven hundred-ten participants (19%) had 2 or more chronic conditions. The first cluster included only myocardial infarction and angina (lift=13.3), indicating that the likelihood of co-occurrence of these conditions is 13 times higher than in isolation. The second cluster consisted of 26 conditions, including cardiovascular, musculoskeletal, respiratory, and neurodegenerative diseases. The strongest association was found between heart failure and atrial fibrillation (lift=23.6). Diabetes was at the center of this cluster with strong associations with heart failure, chronic kidney disease, liver failure, and stroke (lift>2). The third cluster contained 8 highly prevalent conditions, including cancer, hypertension, asthma, and depression, and the strongest association was observed between anxiety and depression (lift=5.0). CONCLUSION: Conditions such as diabetes, hypertension, and asthma are the epicenter of disease clusters for multimorbidity. A more integrative multidisciplinary approach focusing on better management and prevention of these conditions may help prevent other conditions in the clusters.
Authors: Rodrigo Núñez-Cortés; Constanza Malhue-Vidal; Florencia Gath; Gonzalo Valdivia-Lobos; Rodrigo Torres-Castro; Carlos Cruz-Montecinos; Francisco M Martinez-Arnau; Sofía Pérez-Alenda; Rubén López-Bueno; Joaquín Calatayud Journal: Int J Environ Res Public Health Date: 2022-06-18 Impact factor: 4.614
Authors: Yogini V Chudasama; Kamlesh K Khunti; Francesco Zaccardi; Alex V Rowlands; Thomas Yates; Clare L Gillies; Melanie J Davies; Nafeesa N Dhalwani Journal: BMC Med Date: 2019-06-12 Impact factor: 8.775
Authors: Xiaowen Wang; Shanshan Yao; Mengying Wang; Guiying Cao; Zishuo Chen; Ziting Huang; Yao Wu; Ling Han; Beibei Xu; Yonghua Hu Journal: Int J Environ Res Public Health Date: 2020-05-13 Impact factor: 3.390