Year: 2023
Type: article
Abstract: In The Lancet Public Health, Yuntao Chen and colleagues1Chen Y Bandosz P Stoye G et al.Dementia incidence trend in England and Wales, 2002–19, and projection for dementia burden to 2040: analysis of data from the English Longitudinal Study of Ageing.Lancet Public Health. 2023; 8: e859-e867Google Scholar use data from the English Longitudinal Study on Aging (ELSA) to provide evidence for nonlinear trends in dementia incidence between 2002 and 2018. During the beginning of this period (2002–08), dementia incidence decreased, in line with previous studies from the UK and other contexts;2Wolters FJ Chibnik LB Waziry R et al.Twenty-seven-year time trends in dementia incidence in Europe and the United States.Neurology. 2020; 95: e519-e531Crossref PubMed Scopus (0) Google Scholar however, updated analyses show a reversal of this trend starting in 2008. The key findings of Chen and colleagues are robust to sensitivity analyses, including exclusion of administrative data and use of multistate models accounting for missed dementia cases due to interwave mortality. When forecasting models were updated based on new results, estimates of the number of individuals who will have dementia by 2040 were 70% higher than previously reported.1Chen Y Bandosz P Stoye G et al.Dementia incidence trend in England and Wales, 2002–19, and projection for dementia burden to 2040: analysis of data from the English Longitudinal Study of Ageing.Lancet Public Health. 2023; 8: e859-e867Google Scholar The variation in forecasted estimates from prediction models using different underlying assumptions for incidence trends highlights the sensitivity of predictions on required input data. The width of the uncertainty interval reported by Chen and colleagues (0·13 million)1Chen Y Bandosz P Stoye G et al.Dementia incidence trend in England and Wales, 2002–19, and projection for dementia burden to 2040: analysis of data from the English Longitudinal Study of Ageing.Lancet Public Health. 2023; 8: e859-e867Google Scholar is considerably smaller than the width of the analogous uncertainty interval reported in a previous forecasting study (0·59 million for the UK in 2050),3Nichols E Steinmetz JD Vollset SE et al.Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019.Lancet Public Health. 2022; 7: e105-e125Summary Full Text Full Text PDF PubMed Scopus (824) Google Scholar which more explicitly attempted to include uncertainty from various sources. Potential overconfidence in findings as indicated by CIs emphasises the challenges in the use and interpretation of forecasting estimates. Analyses used an algorithmic approach to dementia classification based on DSM-IV criteria, leveraging objective cognitive data, information on self-reported difficulties with daily living activities, and informant reports of cognition when participants were unable to complete cognitive tests. Although physician-given or adjudicated diagnoses are often treated as the gold standard, the authors argue that use of algorithmic approaches can prevent biases due to changes in diagnostic practices over time.4Donegan K Fox N Black N Livingston G Banerjee S Burns A Trends in diagnosis and treatment for people with dementia in the UK from 2005 to 2015: a longitudinal retrospective cohort study.Lancet Public Health. 2017; 2: e149-e156Summary Full Text Full Text PDF PubMed Scopus (94) Google Scholar Algorithmic diagnoses also have increased feasibility in large, population-representative samples. However, algorithmic approaches bring challenges as well. Changes in the cognitive battery over time can pose issues. In ELSA, numeracy questions were only assessed at waves 1, 4, and 6, and verbal fluency was not assessed at wave 6. Evidence is needed to ensure that algorithmic time trends are not affected by these inconsistencies, especially given the brevity of the available battery. As Chen and colleagues note, collection of longitudinal data using more detailed cognitive batteries (eg, the Harmonized Cognitive Assessment Protocol) should yield higher-quality evidence in the future.5Cadar D Abell J Matthews FE et al.Cohort profile update: the Harmonised Cognitive Assessment Protocol sub-study of the English Longitudinal Study of Ageing (ELSA-HCAP).Int J Epidemiol. 2021; 50 (26i): 725Crossref PubMed Scopus (10) Google Scholar Further, the serial application of a cross-sectional algorithm in longitudinal data can lead to reversions, which are conceptually impossible. Chen and colleagues address this problem by removing individuals with decline observed to be transient and applying stringent criteria requiring more severe levels of impairment. The application of stringent criteria increases the specificity of the algorithm and reduces potential bias,6Nichols E Ng DK James BD Deal JA Gross AL The application of cross-sectionally derived dementia algorithms to longitudinal data in risk factor analyses.Ann Epidemiol. 2023; 77: 78-84Crossref Scopus (0) Google Scholar but the use of longitudinal data and information on change over time is likely to improve algorithm performance. The examination of trends in dementia incidence prompts new questions about the drivers of observed changes. Chen and colleagues hypothesise that increases in cases of obesity and type 2 diabetes, as well as improved survival in patients with stroke, could be contributing factors. Although ELSA data show increases in type 2 diabetes over time, the prevalence of BMI appears fairly stable.1Chen Y Bandosz P Stoye G et al.Dementia incidence trend in England and Wales, 2002–19, and projection for dementia burden to 2040: analysis of data from the English Longitudinal Study of Ageing.Lancet Public Health. 2023; 8: e859-e867Google Scholar BMI is an imperfect indicator for obesity,7Adab P Pallan M Whincup PH Is BMI the best measure of obesity?.BMJ. 2018; 360k1274Google Scholar but future work could examine trends in other indicators, such as waist circumference. Chen and colleagues also suggest differences could be due to worsening risk factors in socially disadvantaged groups, supported by subgroup analyses of results stratified by educational attainment. However, alternative explanations for larger increases in groups with lower educational attainment might exist. For example, cohort differences in access to educational opportunities might lead to the presence of fewer individuals with low levels of educational attainment in later years. Subsequent differences in selection processes might lead to differences in the characteristics and composition of the group with lower educational attainment. Despite the limitations of the available evidence, the examination of incidence trends provides an exciting opportunity to better understand the specific risk factors driving observed shifts. The identification of such risks could lead to improved opportunities for public health messaging to motivate both population-level interventions and individual behaviour change aimed at counteracting observed trends. The findings by Chen and colleagues provide a key contribution to the literature by updating previous evidence on trends in dementia incidence and illustrating the subsequent impact of findings on forecasted prevalence. The results emphasise the importance of revisiting existing findings and providing up-to-date estimates using new data. Changes can and should be expected, and new studies are necessary to confirm, update, or challenge previous evidence. Updates are crucial, especially in light of the potential long-term consequences of the COVID-19 pandemic on population-level cognitive health.8Gordon MN Heneka MT Le Page LM et al.Impact of COVID-19 on the onset and progression of Alzheimer's disease and related dementias: a roadmap for future research.Alzheimers Dement. 2022; 18: 1038-1046Crossref Scopus (22) Google Scholar Timely and accurate information is critical for resource allocation, planning efforts for potential interventions, and the organisation of supports and services for patients and their caregivers. The authors declare funding support from the National Institutes of Health: National Institute on Aging (2R01AG030153). Dementia incidence trend in England and Wales, 2002–19, and projection for dementia burden to 2040: analysis of data from the English Longitudinal Study of AgeingDementia incidence might no longer be declining in England and Wales. If the upward trend since 2008 continues, along with population ageing, the burden on health and social care will be large. Full-Text PDF Open Access
Source: The Lancet Public Health
Authors: Emma Nichols, Jinkook Lee
Institution: University of Southern California
Language: English
Subfield: Psychiatry and Mental health
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $5,780
OpenAlex ID: W4387954175
