Frailty progression in adults aged 40 years and older in rural Burkina Faso: a longitudinal, population-based study

Abstract

Background: Little is known about ageing and frailty progression in low-income settings. We aimed to describe frailty changes over time in individuals living in rural Burkina Faso and to assess which sociodemographic, disability, and multimorbidity factors are associated with frailty progression and mortality. Methods: This longitudinal, population-based study was conducted at the Nouna Health and Demographic Surveillance Systems (HDSS) site in northwestern Burkina Faso. Eligible participants were aged 40 years or older and had been primarily resident in a household within the HDSS area for at least the past 6 months before the baseline survey and were selected from the 2015 HDSS household census using a stratified random sample of adults living in unique households within the area. Participants were interviewed in their homes in 2018 (baseline), 2021 (follow-up), or both. We derived the Fried frailty score for each participant at each timepoint using data on grip strength, gait speed, self-reported weight loss, self-reported exhaustion, and physical activity, and described changes in frailty status (no frailty, pre-frailty, or frailty) between 2018 and 2021. We used multivariate regression models to assess factors (ie, sex, age, marital status, educational attainment, wealth quintile, WHO Disability Assessment Schedule (WHODAS) score, and multimorbidity) associated with frailty progression (either worsening frailty status or dying, compared with frailty status remaining the same or improving) and with mortality, and developed sequential models: unadjusted, adjusting for sociodemographic factors (sex, age, marital status, educational attainment, and wealth quintile), and adjusting for sociodemographic factors, disability, and multimorbidity. Findings: Between May 25 and July 19, 2018, and between July 1 and Aug 22, 2021, 5952 individuals were invited to participate: 1709 (28·7%) did not consent, 1054 (17·8%) participated in 2018 only and were lost to follow-up, 1214 (20·4%) participated in 2021 only, and 1975 (33·2%) were included in both years or died between years. Of 1967 participants followed up with complete demographic data, 190 (9·7%) were frail or unable to complete the frailty assessment in 2018, compared with 77 (3·9%) in 2021. Between 2018 and 2021, frailty status improved in 567 (28·8%) participants and worsened in 327 (16·6%), and 101 (5·1%) participants died. The relative risk of frailty status worsening or of dying (compared with frailty impRoving or no change) increased with age and WHODAS score, whereas female sex appeared protective. After controlling for all sociodemographic factors, multimorbidity, and WHODAS score, odds of mortality were 1·07 (odds ratio 2·07, 95% CI 1·05–4·09) times higher among pre-frail individuals and 1·1 (2·21, 0·90–5·41) times higher among frail individuals than among non-frail individuals. Interpretation: Frailty status was highly dynamic in this low-income setting and appears to be modifiable. Given the rapid increase in the numbers of older adults in low-income or middle-income countries, understanding the behaviour of frailty in these settings is of high importance for the development of policies and health systems to ensure the maintenance of health and wellbeing in ageing populations. Future work should focus on designing context-appropriate interventions to improve frailty status. Funding: Alexander Von Humboldt Foundation, Institute for Global Innovation, University of Birmingham, and Wellcome Trust.