Epidemiology of multimorbidity and polypharmacy in ageing: a complementary analysis of mental and brain health in three datasets

Abstract

Multimorbidity, the co-existence of two or more chronic conditions, is common and increasing in prevalence. It is associated with poor outcomes for patients and increased costs for healthcare providers, so is attracting attention both from policymakers and researchers. The use of multiple simultaneous medications (polypharmacy) frequently co-occurs with multimorbidity. Multimorbidity including physical and mental illnesses has been recognised as important and under-studied. It not only poses challenges for patient management but also provides opportunities for interventions which could prevent overall clinical decline. This thesis separates physical and mental illnesses to explore associations between multimorbidity and polypharmacy with mental health outcomes and brain health biomarkers in ageing cohorts. Although there are standard published definitions of multimorbidity, understanding the concept is difficult due to the numerous ways to measure it. This thesis opens with a systematic review of multimorbidity indices. Among 5 560 unique titles identified in a literature search, 35 full-text papers were relevant, and are described and evaluated in detail. Data analysis took place in three datasets focused on ageing, with complementary designs. These are the PREVENT Dementia and European Prevention of Alzheimer’s Dementia (EPAD) study cohorts, and routinely collected data from the National Health Service (NHS) Scotland’s Information Services Division (ISD). In PREVENT Dementia, participants aged 40-59 years are deeply phenotyped, allowing exploration of the epidemiological associations between increasing chronic conditions and medication use with various clinical and biological outcomes. These include self-reported depression, cognitive test results and markers of neurodegeneration on magnetic resonance imaging (MRI). From regression analysis of 210 participants’ data, each additional condition was associated with increased odds of self-reported depression (adjusted OR=1.41, 95% CI 1.11 to 1.80) and anxiety disorder (OR=1.71, 95% CI 1.35 to 2.21). Increasing medication use was associated with self-reported depression (adjusted OR per additional medication=1.36, 95% CI 1.07 to 1.73) but not anxiety disorder (OR=1.24, 95% CI 1.00 to 1.53). There were no meaningful associations between multimorbidity or polypharmacy with MRI or cognitive test outcomes. The EPAD cohort permitted a more focused approach in people aged over 50 years, specifically examining associations between increasing chronic conditions and cerebrospinal fluid (CSF) amyloid-β. In 447 participants, each additional comorbid condition carried a decreased likelihood of amyloid positivity (multiply-adjusted OR=0.82, 95% CI 0.68 to 0.97). This informs the debate that amyloid may not play a part in the pathway between multimorbidity and the development of dementia. Analyses of NHS data used routinely collected information on prescriptions, psychiatric hospital admissions and death certificate diagnoses from 1.23 million people aged over 50 years in Scotland. Adjusted hazard ratios for each additional drug were 1.03 (95% CI 1.03 to 1.04) for death with any psychiatric cause and 1.04 (95% CI 1.04 to 1.05) for admission to psychiatric hospital over 8.5 years of followup. In this and the analyses in PREVENT Dementia, the use of antidepressant or psychotropic medication attenuated the associations. The importance of patient and public involvement in research is also discussed, including perspectives on this work from a Lay Contributor. This thesis explores the measurement of multimorbidity in detail and provides further evidence that physical multimorbidity and polypharmacy are associated with poor mental health. However, the links with biological markers of brain disease such as MRI findings and amyloid are less convincing. This leads to a discussion of possible mechanisms, clinical implications, and proposed future work.