A substantial and growing number of people worldwide are suffering from two or more long-term conditions, known as ‘multimorbidity’. Multimorbidity is associated with functional decline, decreased quality of life, and increased healthcare utilisation. Although there is increasing attention being given to multimorbidity, the management of multiple long-term conditions at the same time is still challenging. The challenge is mainly due to the focus on single conditions in research and clinical guidelines. Thus, there is an urgent need to study multimorbidity to tackle this public health challenge effectively. This thesis aims to characterise the extent and patterns of multimorbidity over time in the UK’s middle-aged and elderly population, investigate the underlying pathways of multimorbidity, and explore possible drug repurposing opportunities to reduce the risks associated with polypharmacy.
The routinely collected UK primary care datasets - Clinical Practice Research Datalink (CPRD) and its linkage datasets and six genome-wide association studies (GWAS) were used. 6,613,198 CPRD participants and 716,733 individuals from GWAS were included in the analyses. Twenty-four chronic conditions, as the Quality and Outcomes Framework (QOF) indicators, were included to study chronic disease co-occurrence. The annual prevalence and incidence of multimorbidity increased from 1987 until 2017. Clustering analyses identified five main clusters, a cardio-metabolic cluster, a neuro-articular cluster, a cardio-mental cluster, a respiratory-cancer cluster, and a cardio-cancer cluster. Disease-specific longitudinal sequences showed osteoarthritis (OA), hypertension, and type II diabetes mellitus (T2DM) were the most common starting conditions, followed by cancer and depression. The real-world observational analysis, based on CPRD data, showed that the overall incidence of Alzheimer's disease (AD) was higher in participants with rheumatoid arthritis (RA), multiple sclerosis (MS), psoriasis, inflammatory bowel disease (IBD) and other inflammatory polyarthropathies & systematic connective tissue disorders (OID). However, the two-sample Mendelian randomisation (MR) analysis, using GWAS data, suggested that genetically predicted risks of inflammatory diseases were not associated with higher AD risk. For potential drug repurposing opportunities, the Cox regression and propensity score weighting (PSW) results suggested an inverse association of methotrexate in reducing the risk of dementia onset in RA patients compared with sulfasalazine. In conclusion, the findings support the increasing trend of multimorbidity in the UK middle-aged and elderly populations, in addition to providing evidence of the associations between inflammatory diseases and AD. The beneficial effects of specific DMARDs on dementia and AD support the hypothesis that diseases may occur under common pathophysiology and provide evidence for future drug discoveries to reduce polypharmacy.