Frailty is an age-related clinical state associated with deterioration across multiple physiological systems and a leading cause of morbidity and mortality later in life. To understand how frailty develops and what causes its progression, longitudinal data with repeated frailty measurements are required. This review summarizes evidence from longitudinal studies on frailty trajectories, transitions, and trends. We identified several consistent findings: frailty increases with aging and is a dynamic condition, and more recent generations of older adults have higher frailty levels. These findings have both clinical and public health relevance, including the provision of healthcare and aged care services in the coming years. Further studies are required, particularly those conducted in low- and middle-income countries and those investigating factors associated with changes in frailty. The latter may help develop better-targeted interventions to reverse or slow the progression of frailty.
The global population of older adults (60 years and over) will nearly double to 2.1 billion people by 2050 because of gains in life expectancy.
The most common condition affecting older adults in terms of both mortality and morbidity is frailty.
Despite the increasing awareness that frailty is preventable, longitudinal data with repeated frailty measurements are still not widely used. This review collated longitudinal studies on frailty to inform the timelines of frailty development and progression in older adults. By examining frailty trajectories, transitions, and trends, we can identify factors that may accelerate or slow frailty development and progression.
Two main conceptual models of frailty exist in the literature today: the phenotypic model first proposed by Fried et al.
The FI model considers frailty to be an accumulation of health deficits, wherein the more deficits individuals have, the frailer they are.
For many years, the field of frailty research has been dominated by studies that evaluate frailty measurements at a single time point. These studies examined the cross-sectional associations between frailty and other variables or used frailty as a predictor of adverse outcomes, particularly mortality.
Various longitudinal studies have examined frailty trajectories using measurements at three or more time points, in which frailty is typically included as a continuous measure (mainly assessed with the FI).
The study of frailty transitions provides another way to examine longitudinal data. To study these transitions, frailty scores must be categorized. Especially for clinical care pathways, the frailty transition approach is helpful because categorization simplifies clinical decision-making, although the classifications may sometimes be arbitrary. Understanding transitions in frailty may help to identify optimal target populations for frailty interventions. Studies have suggested targeting pre-frailty for early interventions.
Studies on frailty trends are relatively scarce as very few longitudinal cohort studies have a cohort-sequential design in which refresher samples of older adults from new generations are added to the study after several years. To our knowledge, only studies in the United States, the Netherlands, the UK, and Sweden have reported on this topic (
Insight into risk factors for certain frailty trajectories, transitions, and trends may inform the development of better-targeted interventions to prevent or reduce frailty levels in older adults. Risk factors are not limited to lifestyle and clinical factors relevant later in life and may also involve behavioral and clinical conditions during early life stages, such as midlife and childhood.
Based on the results of studies on the trajectories, transitions, and trends in frailty among older adults, some implications for clinical practice and public health can be derived. First, the results of the studies discussed in this review emphasize the importance of early identification of frailty. Identification of frailty at an early stage of a frailty trajectory is critical to promote early interventions.
In the past decade, research on trajectories, transitions, and trends in frailty among older adults has emerged. The handful of publications at the beginning of the 2010s has increased in recent years to a substantial research field, with studies using data from large cohorts in various countries. Based on these studies, important clinical implications were derived. For example, frailty prevalence has increased in subsequent generations of older adults, an increase that is expected to continue.
The researchers claim no conflicts of interest.
None.
Conceptualization, EOH, ED; Methodology, EOH; Supervision, EOH; Writing-original draft, EOH,ED; Writing-review & editing, EOH, ED.
Number of articles in PubMed mentioning “frailty trajectories” or “frailty changes” in the past 15 years. The year 2022 was still ongoing when the search was done (November 26, 2022).
Selection of key empirical papers on trajectories, transitions, and trends in frailty among older adults
Study | Year | Data | Frailty instrument | Main findings | |
---|---|---|---|---|---|
Trajectories | Hoogendijk et al. |
2018 | The Netherlands, the Longitudinal Aging Study Amsterdam (LASA), 17-year follow-up | Frailty index (32 items) | Linear increase in frailty with aging. Partner status impacted frailty trajectories, sex and education did not. |
Jenkins et al. |
2022 | Five longitudinal cohorts from the USA, UK, Italy and the Netherlands, 11 to 20 years of follow-up | Frailty index (30–42 items) | Linear increase in frailty over time. No clear association between sex and frailty progression. | |
Raymond et al. |
2020 | Swedish Adoption/Twin Study of Aging (SATSA), 27-year follow-up | Frailty index (42 items) | Frailty increased non-linearly from adulthood into old age. A higher increase was observed after age 65. Deviations from normal weight were associated with frailty trajectories. | |
Stolz et al. |
2017 | 10 European countries, the Survey of Health, Ageing and Retirement in Europe (SHARE), 9-year follow-up | Frailty index (40 items) | Frailty increased with aging (quadratic curve). Different measures of socioeconomic position had a different impact on frailty trajectories. | |
Transitions | Espinoza et al. |
2012 | USA, San Antonio Longitudinal Study of Aging (SALSA), 7-year follow-up | Frailty phenotype | A mix of frailty transitions (worsening, improvement, no change) occurred. A lower level of education was associated with frailty worsening. |
Gill et al. |
2006 | USA, Precipitating Events Project (PEP), 4.5-year follow-up | Frailty phenotype | During follow-up 58% of the study sample had at least one frailty transition. Frailty worsening was more common than frailty improvement. | |
Lee et al. |
2014 | China, longitudinal study, 2-year follow-up | Frailty phenotype | More than half of the sample remained in the same frailty state during follow-up. Various characteristics associated with frailty worsening were identified, such as being male. | |
Romero-Ortuno et al. |
2021 | Ireland, Irish Longitudinal Study on Ageing (TILDA), 8-year follow-up | Frailty phenotype | Frailty was shown to be dynamic. Both frailty worsening (e.g., risk of non-frail to pre-frail progression: 27%) and frailty improvement (e.g., risk of favorable transition from frail to pre-frail: 18%) were common. | |
Trends | Blodgett et al. |
2021 | USA, the National Health and Nutrition Examination Survey (NHANES), 1999–2018 | Frailty index (46 items) | Frailty scores were higher in more recent cohorts in middle-aged (≥35 years) and older adults (≥65 years). Frailty lethality was stable. |
Hoogendijk et al. |
2021 | The Netherlands, the Longitudinal Aging Study Amsterdam (LASA), 1995–2016 | Frailty index (32 items) | The prevalence of frailty increased among people aged 64-84 years in more recent years. The association between frailty and mortality remained the same. | |
Mousa et al. |
2018 | UK, the Cognitive Function and Ageing Studies (CFAS), 1991–2011 | Frailty index (30 items) | Frailty level was higher in 2011 than in 1991 among people aged 65 and over. Its association with mortality did not change. | |
Wennberg et al. |
2022 | Sweden, the 1895–1945 birth cohorts, 1990–2020 | Hospital Frailty Risk Score (HFRS) | Frailty prevalence increased in more recent birth cohorts, at all ages (75, 85 and 95 years). |