2 Adiposity and Development of Dementia: A Systematic Review of Recently Published Longitudinal Studies

Abstract

Abstract Background and Objectives The socioeconomic burden of dementia is very high and on the rise. Currently there is no cure for dementia; therefore, it is key to identify the modifiable risk factors for this condition. The association between adiposity and dementia is not yet convincingly established. This systematic review aimed to critically appraise the quality and to synthesise the findings of recently published longitudinal studies examining the association between adiposity and the development of dementia. Methods A comprehensive literature search on longitudinal studies published between 2013 and 2018 was conducted across MEDLINE, EMBASE, PsycINFO and SCOPUS. The quality of selected papers was assessed using the Newcastle-Ottawa Quality Assessment Scale for cohort studies (NOS) and PRISMA-IPD checklist. The results of the included studies were qualitatively synthesised. Results Of the 1,370 citations identified and reviewed, 10 completely met the inclusion criteria. Of the 10 included studies, 9 were high quality cohort studies, scoring an average of 8.4 out of 9 for NOS. One study was an individual patient data meta-analysis study (IPD). The IPD had high scientific rigour and largely adhered to the PRISMA-IPD checklist. In total, the 9 cohort studies included more than 2 million subjects, with 3 cohorts recruiting exclusively men. The IPD contained data from 39 cohorts and 1.3 million participants. 4 studies were conducted in highly selective cohorts. Apart from European male populations, the other populations were markedly under-represented. All studies used body mass index (BMI) as a proxy for adiposity. None of the included studies did a sample size calculation. Only half of the included studies were able to produce significant results, suggesting insufficient sample size and lack of power. The included studies reached contradictory results, with half of the studies favouring a protective effect of adiposity. Conclusions In view of the results and quality of the studies, this review found that there was insufficient evidence to establish the association between adiposity and dementia. Current evidence suggests an association between mid-life obesity and dementia development in late-life. Longer studies were more likely to conclude that adiposity was detrimental to health. The inconsistent conclusions could be the result of reverse causality. To conclusively establish the link between adiposity and dementia, more research with robust methodology is warranted.