Prevalence and correlates of obesity among the oldest old. A systematic review, meta-analysis and meta-regression

Abstract

Thus far, the prevalence and correlates of obesity among the oldest have been identified in several studies. Nevertheless, there is a lack of a systematic review, meta-analysis and meta-regression synthesizing the existing observational studies. Consequently, our aim was to fill this knowledge gap. Three electronic databases were searched (Medline, PsycINFO, CINAHL) and an additional hand search was performed. Observational studies (both, cross-sectional and longitudinal) determining the prevalence and (preferably) the correlates of obesity in the oldest old (≥80 years) were included. Data extraction covered study design, measurement of obesity, analytical approach, sample characteristics and main results. In total, 19 studies were included. Reported prevalence of obesity in the community varied greatly (range 0.7%–71.5%). The pooled prevalence of obesity was 17.8% (95% CI: 13.3%−22.2%), with significant heterogeneity between studies (I2 = 99.7%, P < 0.001). There is evidence of a publication bias. Meta-regressions showed that some of the heterogeneity was explained by the types of measures of obesity and country of the respective sample. In conclusion, obesity remains a key challenge among the oldest old. Future research in this age bracket is urgently required in regions mostly neglected thus far (e.g., South America, Africa or Asia). Moreover, studies based on longitudinal data are required to clarify the determinants of obesity among the oldest old. Furthermore, studies based on objectively recorded obesity (e.g., waist circumference) are also required. Geriatr Gerontol Int 2022; 22: 373–383.