Back to search

The University of Edinburgh

Dementia and location: lessons from geography and data linkage to routinely collected data

Abstract

dc:description.abstract

Dementia is a growing public health concern, with estimated worldwide prevalence of 152.8 million people by 2050. Dementia prevalence and incidence varies geographically, and this could shed light on risk, and protective, factors for dementia development, or greater understanding if the variation represents differences in accessing a diagnosis for dementia. In this thesis, I use routinely collected data to investigate the association of location - particularly rural and urban living with dementia. In Chapter 1: Introduction, I outline my aims and objectives. I give an overview of dementia and dementia epidemiology in relation to my research question: does dementia vary between rural and urban areas and how can we use routinely collected health data to answer this question? In Chapter 2: Systematic review and meta-analysis of rural or urban living with risk of dementia, I present a systematic review with meta-analysis of observational studies investigating dementia risk with rural versus urban living. Of13,053 records identified, 545 full-text articles were assessed for eligibility, 38 papers were included in the systematic review, of which 27 could be included in a meta-analysis. Overall, rural living was associated with increased odds of dementia, compared to urban living (OR 1.20, 95%CI 1.03-1.40, p=0.018). However, when stratifying by country income, this pattern was seen in Low-Middle Income Countries (L-MICs) (OR 1.33, 1.06-1.68, p=0.014), but in High Income Countries (HICs) there were lower odds of dementia in rural areas compared to urban (OR 1.01, 95% CI 0.88-1.165), p=0.884). Future research should consider heterogeneity in the definitions of ‘rural’ versus ‘urban’, as well as growing disparity in methodologies between HIC (large routinely collected data studies) and L-MIC (traditional smaller face-to-face epidemiological studies). In Chapter 3: Using big data for dementia research, I give an overview of using routinely collected healthcare data (often referred to as ‘big data’) for research, and how these could be used for dementia research. I present my peer-reviewed article on big data in Scotland, and also discuss the health data landscape across the UK. Finally, I consider how future ageing health research using routinely collected healthcare data could expand in the UK. In the following Chapters (4, 5, and 6) I build on the preceding work to present novel research in Aotearoa New Zealand and Scotland using routinely collected data to investigate variation in location and dementia. In Chapter 4: Investigating dementia in Aotearoa New Zealand (ANZ) using routinely collected data, I investigate the variation of dementia rates between North and South Island in ANZ using a large national data resource, the Integrated Data Infrastructure (IDI), comprising of many routinely collected datasets. The prevalence of dementia in the ≥ 60 year olds varied including between North and South Island (e.g., for 2012, North = 9,696/637,245 people [1.52%], and South = 3,702/230,001 people [1.61%]). Dementia counts were highest amongst the European ethnicity but when age and sex standardised, dementia cases were higher for Māori than non-Māori (2% vs.1%). My methods have since been used by local and international research collaborations investigating dementia in ANZ, and additionally work from other research teams have since linked additional datasets to the IDI to increase dementia ascertainment. In Chapter 5: Methodology for the SMS1947 dementia data linkage study, I outline the processes which were necessary to undertake a research project investigating dementia variation in a historical population level survey, the Scottish Mental Survey 1947 (SMS1947), with individual-level data CHI-linked to routinely collected health data, focusing on investigating rural and urban variation and dementia diagnoses. This chapter highlights the stages of project development, health data linkage methodology, and the challenges encountered. It outlines the methodology used for Chapter 6: Results of the SMS1947 dementia data linkage study. In Chapter 6: Results of the SMS1947 dementia data linkage study, I investigate dementia within the Scottish Mental Survey 1947 (SMS1947). Living in a rural region was associated with lower hazard of dementia: HR adjusted for sex, Intelligence score, and Scottish Index of Multiple Deprivation (SIMD) 0.84, 95% CI 0.77-0.93 compared to living in an urban area. Dementia was even lower if living within the most rural/remote region of Scotland (adjusted HR 0.77, 95% CI 0.61-0.98). A major strength of this study was the ability to link a historical national survey (which has a childhood intelligence variable, allowing adjustment for this potential confounder) to healthcare datasets, thus creating a population-level cohort with individual level data from 1947. In the final Chapter 7: Discussion I summarise my findings and discuss the impact and implications of my research alongside the strength and limitations. I also outline, and discuss, future research needed and then conclude with final remarks with a focus on the changing landscape of using routinely collected health data as a secondary use in medical research.

Degree

thesis:*
Grantor dc:publisher
The University of Edinburgh
Year dc:date.issued
2025

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Walesby, Katherine Elizabeth
Advisors dc:contributor.advisor
  • Russ, Tom
  • Shenkin, Susan

Subjects

dc:subject × 5

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
University of Edinburgh
Base URL
era.ed.ac.uk/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Walesby, Katherine Elizabeth. Dementia and location: lessons from geography and data linkage to routinely collected data. The University of Edinburgh, 2025. https://hdl.handle.net/1842/43512