Abstract
Background
Falls and frailty are common, interrelated conditions and major contributors to disability and mortality. While frailty is a recognized risk factor for falls, less is known about whether falls influence subsequent frailty transitions. We examined the association between falls and transitions across frailty states and mortality in community-dwelling older adults.
Methods
We analyzed longitudinal data from the China Health and Retirement Longitudinal Study (2011–2020). Participants aged ≥60 years with at least two survey waves were included. Frailty was assessed using a 32-item frailty index and categorised as robust, pre-frail, frail, and death (absorbing state). Falls were treated as a time-varying exposure. Frailty transitions were modelled using a multi-state survival framework with transition-specific incidence rate ratios (IRRs), adjusting for sociodemographic factors, health behaviors and pain.
Findings
Among 12,501 participants (44,129 observations), falls were associated with adverse frailty transitions. Falls were associated with increased progression from robust to pre-frail (IRR 1.18, 95% CI 1.09–1.27) and pre-frail to frail (1.51, 1.42–1.60), reduced recovery, and higher mortality from pre-frail and frail states. Similar but generally stronger associations were observed for injurious falls. Women accumulated frailty more rapidly over time, whereas men experienced higher mortality, particularly following a fall. Chronic pain, older age, and socioeconomic disadvantage further accelerated frailty progression.
Interpretation
Falls represent clinically actionable sentinel events in the frailty life-course, accelerating transitions toward disability and long-term care needs. Integrating falls prevention and post-fall rehabilitation within frailty care pathways may slow functional decline and improve healthy ageing outcomes.
Keywords: Key words, Frailty, Falls, Multistate models, Ageing, Longitudinal studies
1. Introduction
Accelerated demographic ageing in the Asia-Pacific region reflects sustained improvements in life expectancy and reductions in fertility across countries including China, Japan, Singapore, and the Republic of Korea, leading to a rising proportion of older adults and widening gaps between lifespan and healthspan [1]. In China, the proportion of adults aged ≥65 years is projected to increase from 14.9% in 2025 to approximately 30·9% by 2050, consistent with a transition towards a super-aged demographic profile [1]. These demographic changes are associated with a rising prevalence of non-communicable diseases, frailty, dementia, and disability, which can pose major challenges to older adults’ health and quality of life.
Falls are a leading cause of injury, disability, and mortality among older adults worldwide, with prevalence estimates among community-dwelling older people ranging from approximately 14 % to 30 % in several Asia-Pacific settings and as high as 53 % in some populations [2,3]. Beyond injury, falls are associated with fear of falling (FOF), reduced social participation, poor perceived health, frailty, sarcopenia, and depression [4]. In the United States, annual medical costs attributable to fatal and non-fatal falls exceeded US$50 billion in 2015, with non-fatal fall spending estimated to approach US$80 billion in recent years [5]. Importantly, falls are preventable, with evidence supporting the effectiveness of exercise-based and multifactorial interventions in reducing fall incidence among community-dwelling older adults [6,7].
Frailty is a multidimensional geriatric syndrome characterized by declining physiological reserve and increased vulnerability to stressors and adverse health outcomes [8]. Frailty and pre-frailty are common dynamic conditions in older adults with prevalence of pre-frailty ranging between 13% and 49% and frailty between 4% and 49% depending on the population studied and screening tools used [[9], [10], [11], [12]]. Commonly used frailty assessment tools include Clinical Frailty Scale, Frailty Phenotype, FRAIL scale or Frailty Index [8]. The World Falls Guidelines recommend multifactorial falls risk assessment incorporating frailty and sarcopenia [7]. Falls and frailty share a bidirectional relationship mediated by overlapping risk factors such as malnutrition, sarcopenia, polypharmacy, depression, cognitive impairment, multimorbidity, and exposure to falls-risk-increasing drugs [13].
Although frailty is well recognized as a risk factor for falls [14], less attention has been given to falls as a potential driver of frailty transitions. This evidence gap is reflected in clinical practice, where frailty prevention programmes and guidelines seldom integrate falls prevention within frailty care models [8,15,16]. Similarly, many clinical guidelines on frailty rarely emphasize on upstream prevention of falls [8,15]. Existing studies have largely examined cross-sectional associations between frailty and falls, with relatively few investigating how incident falls influence transitions between frailty states over time [17]. To address this gap, the present study examines how incident falls influence transitions between robust, pre-frail, and frail states, as well as mortality.
2. Methods
2.1. Study design and data source
This study is a secondary analysis of longitudinal data from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative cohort of community-dwelling adults aged 45 years and older. CHARLS employed a stratified, multistage, probability-proportional-to-size sampling design across 28 provinces in China [18]. The baseline survey in 2011–2012 enrolled 17,708 individuals, with follow-up waves in 2013, 2015, 2018, and 2020. Response rates exceeded 80% across waves [19]. To maintain representativeness, CHARLS also introduced refreshment samples at subsequent waves.
For the present analysis, participants aged ≥60 years with observations from at least two survey waves between 2011 and 2020 were included, to allow observation of transitions in frailty status (see Supplementary Fig. S1 for participant flow). Participants were censored at their last available follow-up or at the time of death. Deaths during follow-up were ascertained through exit interviews with family members. Participant who died were assigned to the “death” state for transition analysis, with the year of death recorded, or imputed as the survey year when death date was unavailable.
2.2. Frailty state classification
Frailty was measured using a 32-item Frailty Index (FI) based on a cumulative deficit model [20,21]. The FI included deficits covering self-rated health, chronic diseases and neuropsychiatric conditions, mobility limitations, impairment in activities of daily living (ADL) and instrumental ADL (IADL), sensory impairments, depressive symptoms, and cognitive function (Supplementary Table S1). Each deficit was coded as present (1) or absent (0); cognitive function was treated as a continuous score ranging from 0 to 1 derived from a combined memory and mental status test. Participants with more than 10% missing FI items were excluded. Seven mobility limitation items not collected in the 2020 wave were imputed using multiple imputation by chained equations to ensure consistency across waves. FI scores ranged from 0 to 1, with higher values indicating greater frailty.
The FI was categorised into “robust” (≤0·08), “pre-frail” (>0·08 to <0·25), and “frail” (≥0·25) [21] with death included as an absorbing state. Participants therefore occupied one of four mutually exclusive states (robust, pre-frail, frail, or death) at each wave, allowing both improvement and worsening in frailty over time, as well as mortality. Transitions to the death state could occur from any of the three living states, with no further transitions thereafter.
2.3. Falls measurement
Falls were treated as a time-varying exposure. At each follow-up, participants reported whether they had experienced any fall since the previous interview (approximately 2–3-year interval), coded dichotomously for analysis [22]. Additionally, injurious falls were coded as an alternative time-varying exposure, indicating whether the participant had experienced any fall requiring medical treatment during the same interval (Supplementary Table S2).
2.4. Covariates
We adjusted for a range of sociodemographic, behavioral, and health-related factors that might confound the relationship between falls and frailty transitions. Sociodemographic variables comprised age group (60–69, 70–79, or ≥ 80 years), sex, marital status (married or cohabiting vs. unmarried/ divorced/ separated/ widowed), education level (illiterate or below primary school, primary school, or middle school and above), place of residence (rural vs. urban), and geographic region (east, central, west, or northeast China). Behavioral covariates included smoking and alcohol use (never, former, and current). Chronic pain was coded dichotomously and defined as self-reported frequent bodily pain. Age, marital status, smoking status, drinking status, and pain status were treated as time-varying covariates. Missing covariate data were imputed using adjacent waves; remaining unresolved missingness was excluded.
2.5. Statistical analysis
Frailty states were assigned at each observed wave (2011–2020). Between consecutive waves, participants could remain in the same state or transition to a different state. Baseline characteristics were summarized as frequencies and compared across different frailty states with Pearson’s Chi-square tests.
Frailty transitions between successive observations were modelled using a multi-state survival framework with piecewise-constant hazards, implemented through Poisson regression with a log link and an offset for the person-time at risk. This approach is mathematically equivalent to a continuous-time Markov model observed at discrete follow-up waves and accommodates unequal interval lengths and time-varying covariates. Possible seven transitions included recovery (frail to pre-frail; pre-frail to robust), progression (robust to pre-frail; pre-frail to frail), and death from any living state.
Data were structured at the person-interval level, with each interval contributing to the risk set for all transitions originating from the starting state. Transition-specific hazards were estimated for each allowable transition. Falls and covariates were included simultaneously, allowing transition-specific effect estimation. Interval length was incorporated as an offset to account for differing durations of time at risk. We calculated robust standard errors clustered on the individual level to account for within-person correlation across repeated observations. Sensitivity analysis was conducted restricting follow-up to the first four waves (excluding 2020) to assess potential bias related to the COVID-19 pandemic. Supplementary analysis was conducted using injurious falls as a more clinically severe fall definition. Effect estimates are reported as incidence rate ratios (IRRs) with 95% confidence intervals (CIs). All analyses were performed using R (version 4.4.1). Statistical significance was defined as a two-tailed P-value less than 0.05.
3. Results
A total of 12 501 participants aged ≥60 years contributed 44 129 observations between 2011 and 2020. At baseline, 32.2% were classified as robust, 45.0% as pre-frail, and 22.8% as frail (Table 1). Frail participants were older, more often female, unmarried, less educated, and more likely to reside in rural areas and western regions of China. Chronic pain was common and increased across frailty states (robust 16.6%, pre-frail 41.0%, frail 69.0%). Overall, 19.2% participants reported ≥1 fall and 9.9% reported ≥1 injurious fall, higher among frail individuals (32.5% for fall and 18.3% for injurious fall) than among robust individuals.
Table 1.
Baseline characteristics of all participants.
| Baseline characteristics | Overalla (N = 12,501) | Frailty statea |
P-valueb | ||
|---|---|---|---|---|---|
| Robust (N = 4028) | Pre-frail (N = 5624) | Frail (N = 2849) | |||
| Entry wave | <0.001 | ||||
| 2011 | 6626 (53.0%) | 2022 (50.2%) | 2988 (53.1%) | 1616 (56.7%) | |
| 2013 | 2087 (16.7%) | 776 (19.3%) | 907 (16.1%) | 404 (14.2%) | |
| 2015 | 1561 (12.5%) | 549 (13.6%) | 694 (12.3%) | 318 (11.2%) | |
| 2018 | 2227 (17.8%) | 681 (16.9%) | 1035 (18.4%) | 511 (17.9%) | |
| Have fallen | 2395 (19.2%) | 434 (10.8%) | 1036 (18.4%) | 925 (32.5%) | <0.001 |
| Have fallen injuriously | 1231 (9.9%) | 217 (5.4%) | 493 (8.8%) | 521 (18.3%) | <0.001 |
| Age group | <0.001 | ||||
| 60-69 | 9475 (75.8%) | 3294 (81.8%) | 4326 (76.9%) | 1855 (65.1%) | |
| 70–79 | 2450 (19.6%) | 645 (16.0%) | 1086 (19.3%) | 719 (25.2%) | |
| 80– | 576 (4.6%) | 89 (2.2%) | 212 (3.8%) | 275 (9.7%) | |
| Gender | <0.001 | ||||
| Female | 6272 (50.2%) | 1607 (39.9%) | 2887 (51.3%) | 1778 (62.4%) | |
| Male | 6229 (49.8%) | 2421 (60.1%) | 2737 (48.7%) | 1071 (37.6%) | |
| Married/Partnered | 9900 (79.2%) | 3331 (82.7%) | 4487 (79.8%) | 2082 (73.1%) | <0.001 |
| Education attainment | <0.001 | ||||
| Below primary school | 6908 (55.3%) | 1804 (44.8%) | 3120 (55.5%) | 1984 (69.6%) | |
| Primary school | 2656 (21.2%) | 973 (24.2%) | 1208 (21.5%) | 475 (16.7%) | |
| Middle school and above | 2937 (23.5%) | 1251 (31.1%) | 1296 (23.0%) | 390 (13.7%) | |
| Rural residence | 7525 (60.2%) | 2202 (54.7%) | 3374 (60.0%) | 1949 (68.4%) | <0.001 |
| Region | <0.001 | ||||
| East | 3923 (31.4%) | 1489 (37.0%) | 1748 (31.1%) | 686 (24.1%) | |
| Central | 3589 (28.7%) | 1120 (27.8%) | 1652 (29.4%) | 817 (28.7%) | |
| West | 4145 (33.2%) | 1160 (28.8%) | 1849 (32.9%) | 1136 (39.9%) | |
| Northeast | 844 (6.8%) | 259 (6.4%) | 375 (6.7%) | 210 (7.4%) | |
| Chronic pain | 4938 (39.5%) | 667 (16.6%) | 2306 (41.0%) | 1965 (69.0%) | <0.001 |
| Smoke status | <0.001 | ||||
| Never | 7052 (56.4%) | 2060 (51.1%) | 3186 (56.7%) | 1806 (63.4%) | |
| Current smoker | 3956 (31.6%) | 1554 (38.6%) | 1712 (30.4%) | 690 (24.2%) | |
| Former smoker | 1493 (11.9%) | 414 (10.3%) | 726 (12.9%) | 353 (12.4%) | |
| Drink status | <0.001 | ||||
| Never | 8217 (65.7%) | 2449 (60.8%) | 3717 (66.1%) | 2051 (72.0%) | |
| Current drinker | 3156 (25.2%) | 1285 (31.9%) | 1404 (25.0%) | 467 (16.4%) | |
| Former drinker | 1128 (9.0%) | 294 (7.3%) | 503 (8.9%) | 331 (11.6%) | |
n (%).
Pearson's Chi-squared Test.
Falls were associated with increased frailty progression, reduced recovery, and higher mortality (Fig. 1, Table S3). Among robust older adults, falls increased transitions to pre-frail (IRR 1.18, 95% CI 1.09–1.27). Among pre-frail adults, falls substantially increased progression to frailty (IRR 1.51, 95% CI 1.42–1.60) and reduced recovery to robust status (IRR 0.49, 95% CI 0.42–0.58). Among frail adults, falls reduced recovery to pre-frail states (IRR 0.53, 95% CI 0.48–0.59). Falls were also associated with increased mortality from pre-frail (IRR 1.41, 95% CI 1.07–1.87) and frail states (IRR 1.33, 95% CI 1.11–1.58), whereas associations with mortality from the robust state were weak.
Fig. 1.
Forest plots of covariate effects on transitions between frailty states.
*Among adults aged ≥80 years, mortality-related IRR estimates (range 5.45–5.65) exceeded the upper limit of the plotting range and are not displayed.
Several individual and contextual determinants were independently associated with transitions (Fig. 1, Table S3). Chronic pain emerged as one of the strongest predictors of deterioration, elevating the risks of transitioning from robust to pre-frail (IRR 1.34, 95% CI 1.26–1.43) and from pre-frail to frail (IRR 1.40, 95% CI 1.33–1.48), while significantly reducing the likelihood of recovery from pre-frail to robust (IRR 0.72, 95% CI 0.65–0.79). Older age was associated with increased deterioration and substantially higher mortality across all starting states, with adults aged 80 years or older experiencing more than a fivefold increase in mortality. Sex differences followed a mixed pattern, whereby men were less likely to progress from robust to pre-frail (IRR 0.84, 95% CI 0.77–0.92) but faced much higher mortality when already frail (up to IRR 2.47, 95% CI 1.95–3.13). Higher educational attainment appeared protective, as individuals with at least a middle-school education had a lower risk of transitioning from pre-frail to frail (IRR 0.70, 95% CI 0.64–0.77). Rural residence modestly increased the probability of worsening transitions from pre-frail to frail (IRR 1.20, 95% CI 1.12–1.27). Compared with older adults residing in the more economically developed eastern region of China, those living in central, western, and northeastern regions were more likely to transition from pre-frail to frail (IRR range 1.15–1.33) and were less likely to recover from frail to pre-frail (IRR range 0.79–0.87). Former smokers and former drinkers were also more likely to experience worsening transitions and less likely to recover to healthier states than never users. Sensitivity analyses restricted to the first four waves yielded results consistent with the primary findings (Table S4).
Fig. 2 illustrates model-based state occupancy probabilities for the cohort over follow-up. Predicted trajectories showed that participants with falls had progressively lower probabilities of remaining robust and higher probabilities of occupying frail or death states than those without falls. These divergences widened over time, indicating cumulative disadvantage following falls (Table S6).
Fig. 2.
Stacked predicted state occupancy probabilities with time, by gender and fall history.
*Transition probability matrices and time-dependent state occupancy probabilities were derived from the estimated transition intensity matrix using matrix exponential methods.
Sex-specific patterns were evident. Women, regardless of fall history, exhibited a faster increase in the probability of occupying pre-frail and frail states over time, whereas men showed higher mortality risks. Among participants with prior falls, these sex differences became more pronounced: women experienced a steeper decline in the probability of remaining robust and a more rapid rise in frailty prevalence across the six-year horizon, whereas men showed a comparatively slower shift toward frailty but a greater increase in the likelihood of death.
Supplementary analysis using injurious falls showed a pattern consistent with, and generally stronger than the primary analysis based on any fall, especially among initially robust older adults (Table S5). Injurious falls were associated with increased progression from robust to pre-frail (IRR 1.25, 95% CI 1.13–1.38) and from pre-frail to frail status (IRR 1.48, 95% CI 1.38–1.60), reduced recovery from pre-frail to robust (IRR 0.49, 95% CI 0.38–0.63) and from frail to pre-frail status (IRR 0.65, 95% CI 0.56–0.76), and higher mortality from robust, pre-frail, and frail states. Model-based transition probabilities stratified by sex and injurious fall history showed patterns consistent with the primary fall analysis, with injurious falls associated with lower probabilities of remaining robust and higher probabilities of frailty and death over time (Table S7, Fig. S2).
4. Discussion
In this longitudinal multi-state analysis, falls were associated with adverse transitions in frailty status, including progression from robust to pre-frail or frail states and increased mortality. While most evidence has conceptualized frailty as a risk factor for falls, our findings extend this paradigm by demonstrating that falls may actively accelerate frailty transitions and reduce the likelihood of recovery [23]. These findings suggest that falls may represent clinically relevant sentinel events within the frailty life course.
Baseline gradients in frailty status by age, sex, and socioeconomic position were consistent with patterns reported in Chinese community cohorts. In a systematic review, Zhou et al. found frailty prevalence of 10.1% and pre-frailty 43.9% among community-dwelling older adults, with higher prevalence in those aged ≥70 years, women, unmarried individuals, and those with lower education, rural residence, and poorer socioeconomic conditions [24]. Transitions from any frailty state to death increased markedly with advancing age, exceeding twofold among those aged ≥70 years and more than fivefold among those aged ≥80 years. This age gradient is biologically plausible, reflecting declining physiological reserve, impaired recovery capacity, and a higher prevalence of conditions such as osteoporosis that increase susceptibility to fall-related injury and complications [25]. Sex differences in frailty transitions and mortality observed in this study align with prior evidence [26].Women showed more rapid frailty accumulation over time, whereas men experienced higher mortality, a pattern likely reflecting combined biological, behavioral, and social factors. Biological differences in muscle–fat composition, hormonal profiles, osteoporosis burden, and longer female survival may predispose women to prolonged frailty trajectories, while men may transition more directly from frailty to death. Sex-specific health behaviors, including greater fall reporting and healthcare utilization among women [27], may further shape divergent frailty pathways, particularly following a fall.
In our study, chronic pain was independently associated with accelerated transitions along the frailty spectrum, consistent with evidence linking persistent pain to impaired mobility, balance deficits, psychological distress, and reduced physical activity [28,29]. A 2025 longitudinal study using data from the China Health and Retirement Longitudinal Study, the English Longitudinal Study of Ageing, and the US Health and Retirement Study demonstrated a bidirectional relationship between pain and frailty. Prefrail individuals had nearly a five-fold increased risk of pain, while those with frailty had almost a ten-fold higher risk [30]. Both smoking and alcohol consumption are associated with reduced bone mineral density, increased osteoporosis risk, and impaired muscle function, which may heighten susceptibility to fall-related injury and subsequent frailty progression [31]. Former smokers and alcohol drinkers demonstrated elevated risks across multiple frailty transitions. This pattern is consistent with the “sick quitter” effect, where individuals cease smoking or alcohol use due to declining health, producing reverse causation and higher observed risk among former users [32]. Together, these findings highlight the complex interplay between pain, health behaviors, and biological vulnerability in shaping frailty trajectories following falls.
The frailty index was used to measure frailty on our study. According to geroscience frameworks, biological ageing processes begin before clinically apparent declines in function [33]. Subclinical changes such as inflammation, mitochondrial dysfunction, and reduced stress-response capacity may increase vulnerability to falls and partially explain why falls can occur before measurable frailty progression [34]. Once a fall occurs, downstream consequences including injury, pain, activity restriction, cognitive impairment, and deconditioning may accelerate frailty trajectories [35]. These consequences can collectively account towards a loss of physiological reserve and reinforcing adverse frailty trajectories.
Psychological sequelae of falls such as FoF may also influence frailty trajectories, with evidence supporting a bidirectional relationship between FoF and frailty [[36], [37], [38]]. Although FoF was not captured in this study, a longitudinal Korean cohort showed that activity limitations due to FoF accelerated frailty progression [39] and findings from Irish and Malaysian ageing cohorts linked adverse falling trajectories with poorer cognitive function, FoF, antidepressant use, and declines in physical function [40,41]. Fall-related psychological factors may therefore represent targets for intervention.
5. Clinical and public health implications
Frailty and falls are modifiable, and exercise-based interventions can delay frailty progression and reduce fall risk among community-dwelling older adults. However, higher risk of injurious falls in frailty highlight the need to balance activity with risk mitigation and rehabilitation that enhances confidence and falls efficacy [[42], [43], [44]].
At the population level, integrating falls prevention and post-fall intervention within frailty management programmes may help delay deterioration and reduce mortality. Global guidelines emphasize multidomain interventions for both falls prevention and frailty management, but these require tailoring to individual risk profiles and functional priorities, such as balance, strength, gait, endurance, and psychological confidence. The observed urban–rural and regional disparities further highlight the need for public health responses that address nutritional vulnerability, education, environmental hazards, and socioeconomic disadvantage. Policy frameworks for healthy ageing should consider falls as an early and perceptible phenotype that can trigger timely assessment and intervention for frailty, given that frailty often requires more systematic clinical assessment to detect.
Collectively, these implications emphasize the value of integrated models that combine falls prevention, frailty screening, and rehabilitation approaches to support healthy ageing across the life course.
6. Limitations
This study is strengthened by using a large, nationally representative longitudinal cohort with repeated frailty assessments and a multi-state modelling framework to characterize dynamic frailty transitions. Nonetheless, several limitations should be noted. First, falls were self-reported over a 2–3-years recall period and modelled as interval-based exposures, without information on timing, recurrence or injury severity, which may have limited temporal precision. Second, although falls were treated as preceding subsequent frailty transition, reverse causation cannot be fully excluded given the inherently bidirectional relationship between frailty and fall risk. Third, CHARLS did not collect sufficiently detailed information on Parkinson disease, psychotropic medication use, or other fall-risk-increasing drugs across waves, residual confounding therefore cannot be excluded. In addition, the 32-item frailty index included mobility-related items that may partially overlap with fall-related deficits, potentially inflating associations. Finally, several mobility limitation items required to construct the frailty index were not collected in the 2020 wave and were imputed to ensure consistency across waves; sensitivity analyses excluding this wave yielded similar results, although this wave-specific missingness remains a methodological consideration.
7. Conclusion
In this large national longitudinal cohort, falls were associated with adverse progression across frailty states and increased mortality. These findings position falls not merely as a consequence of frailty, but as a potentially modifiable turning point in the frailty life course. Framed within a dynamic frailty framework, falls emerge as clinically actionable sentinel events that signal heightened vulnerability and provide a critical opportunity for integrated prevention and post-fall interventions to slow frailty progression and support healthy ageing at the population level.
CRediT authorship contribution statement
YL conducted the data analysis and drafted the manuscript. Shawn LH Soh and Reshma Aziz Merchant contributed to study conceptualization, methodology, interpretation of results, and manuscript revision. Maw Pin Tan, Wenjing Ji, Wayne LS Chan, Devinder Kaur Singh, R Boonsinsukh, Christopher TC Lien, Ryuichi Sawa, Susiana Nugraha, Shahrimawati Binti Hj Sharbini, Shyh Poh Teo, Shehan Silva, and Jagadish K. Chhetri contributed to interpretation of findings and critical revision of the manuscript.
All authors saw and approved the final version, and no other person made a substantial contribution to the paper.
Ethical statement
The China Health and Retirement Longitudinal Study (CHARLS) was approved by the Institutional Review Board at Peking University (approval number: IRB00001052-11015). Written informed consent was obtained from all CHARLS participants before data collection. The present study used publicly available, de-identified secondary data and did not require additional ethics approval.
Declaration of Generative AI and AI-assisted technologies in the writing process
Generative AI tools were used only to assist with language editing and text refinement. All scientific content, analysis, interpretation, and conclusions were developed and verified by the authors.
Funding
This research received no grant.
Data statement
The data used in this study are publicly available from the China Health and Retirement Longitudinal Study website (https://charls.pku.edu.cn/en/) upon registration and application.
Declaration of competing interest
None declared.
Acknowledgements
The authors thank the CHARLS group and all participants for providing the survey data used in this study.
Footnotes
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jnha.2026.100908.
Appendix A. Supplementary data
The following is Supplementary data to this article:
References
- 1.UN.ESCAP . United Nations, Economic and Social Commission for Asia and the Pacific; 2025. ESCAP population data sheet 2025. [Google Scholar]
- 2.Salari N., Darvishi N., Ahmadipanah M., Shohaimi S., Mohammadi M. Global prevalence of falls in the older adults: a comprehensive systematic review and meta-analysis. J Orthop Surg Res. 2022;17(1):334. doi: 10.1186/s13018-022-03222-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Montero-Odasso M., Pieruccini-Faria F., Son S., Carvalho de Abreu D.C., Hunter S., Liu J.Q., et al. Fall risk stratification in older adults: low and not-at-risk status still associated with falls and injuries. Age Ageing. 2025;54(3) doi: 10.1093/ageing/afaf064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Thomas S.M., Parker A., Fortune J., Mitchell G., Hezam A., Jiang Y., et al. Global evidence on falls and subsequent social isolation in older adults: a scoping review. BMJ Open. 2022;12(9) doi: 10.1136/bmjopen-2022-062124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Haddad Y.K., Miller G.F., Kakara R., Florence C., Bergen G., Burns E.R., et al. Healthcare spending for non-fatal falls among older adults, USA. Inj Prev. 2024;30(4):272–276. doi: 10.1136/ip-2023-045023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pillay J., Gaudet L.A., Saba S., Vandermeer B., Ashiq A.R., Wingert A., et al. Falls prevention interventions for community-dwelling older adults: systematic review and meta-analysis of benefits, harms, and patient values and preferences. Syst Rev. 2024;13(1):289. doi: 10.1186/s13643-024-02681-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Montero-Odasso M., van der Velde N., Martin F.C., Petrovic M., Tan M.P., Ryg J., et al. World guidelines for falls prevention and management for older adults: a global initiative. Age Ageing. 2022;51(9) doi: 10.1093/ageing/afac205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dent E., Morley J.E., Cruz-Jentoft A.J., Woodhouse L., Rodriguez-Manas L., Fried L.P., et al. Physical frailty: ICFSR international clinical practice guidelines for identification and management. J Nutr Health Aging. 2019;23(9):771–787. doi: 10.1007/s12603-019-1273-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Davidson S.L., Lee J., Emmence L., Bickerstaff E., Rayers G., Davidson E., et al. Systematic review and meta-analysis of the prevalence of frailty and pre-frailty amongst older hospital inpatients in low- and middle-income countries. Age Ageing. 2025;54(1) doi: 10.1093/ageing/afae279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.O’Caoimh R., Sezgin D., O’Donovan M.R., Molloy D.W., Clegg A., Rockwood K., et al. Prevalence of frailty in 62 countries across the world: a systematic review and meta-analysis of population-level studies. Age Ageing. 2020;50(1):96–104. doi: 10.1093/ageing/afaa219. [DOI] [PubMed] [Google Scholar]
- 11.Merchant R.A., Chen M.Z., Tan L.W.L., Lim M.Y., Ho H.K., van Dam R.M. Singapore Healthy Older People Everyday (HOPE) study: prevalence of frailty and associated factors in older adults. J Am Med Dir Assoc. 2017;18(8):734.e9–734.e14. doi: 10.1016/j.jamda.2017.04.020. [DOI] [PubMed] [Google Scholar]
- 12.Siriwardhana D.D., Hardoon S., Rait G., Weerasinghe M.C., Walters K.R. Prevalence of frailty and prefrailty among community-dwelling older adults in low-income and middle-income countries: a systematic review and meta-analysis. BMJ Open. 2018;8(3) doi: 10.1136/bmjopen-2017-018195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kojima G. Frailty as a predictor of future falls among community-dwelling older people: a systematic review and meta-analysis. J Am Med Dir Assoc. 2015;16(12):1027–1033. doi: 10.1016/j.jamda.2015.06.018. [DOI] [PubMed] [Google Scholar]
- 14.Yang Z.C., Lin H., Jiang G.H., Chu Y.H., Gao J.H., Tong Z.J., et al. Frailty is a risk factor for falls in the older adults: a systematic review and meta-analysis. J Nutr Health Aging. 2023;27(6):487–495. doi: 10.1007/s12603-023-1935-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mehta P., Lemon G., Hight L., Allan A., Li C., Pandher S.K., et al. A Systematic Review of Clinical Practice Guidelines for Identification and Management of Frailty. The Journal of nutrition, health and aging. 2021;25(3):382–391. doi: 10.1007/s12603-020-1549-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rasiah J., Prorok J.C., Adekpedjou R., Barrie C., Basualdo C., Burns R., et al. Enabling healthy aging to AVOID frailty in community dwelling older Canadians. Can Geriatr J. 2022;25(2):202–211. doi: 10.5770/cgj.25.536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kim Y.-S., Yao Y., Lee S.-W., Veronese N., Ma S.-J., Park Y.-H., et al. Association of frailty with fall events in older adults: A 12-year longitudinal study in Korea. Arch Gerontol Geriatr. 2022;102 doi: 10.1016/j.archger.2022.104747. [DOI] [PubMed] [Google Scholar]
- 18.Zhao Y., Hu Y., Smith J.P., Strauss J., Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS) Int J Epidemiol. 2014;43(1):61–68. doi: 10.1093/ije/dys203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhao Y., Strauss J., Chen X., Wang Y., Gong J., Meng Q., et al. National School of Development, Peking University: National School of Development, Peking University; 2020. China health and retirement longitudinal study wave 4 user’s guide. [Google Scholar]
- 20.Tong Y., Teng Y., Zhang Y., Huang C., Liao W., Wan B., et al. Bidirectional transitions of frailty states among middle-aged and older adults: a longitudinal cohort analysis using a multi-state Markov model based on the China Health and Retirement Longitudinal Study. Innov Aging. 2025;9(12) doi: 10.1093/geroni/igaf095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Romero-Ortuno R. An alternative method for Frailty Index cut-off points to define frailty categories. Eur Geriatr Med. 2013;4(5) doi: 10.1016/j.eurger.2013.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Liu Y., Kabba J.A., Xu S., Gu H., Su X., Liu Y., et al. Regional and temporal trends of falls and injurious falls among Chinese older adults: results from China Health and Retirement Longitudinal Study, 2011-2018. Inj Prev. 2023;29(5):389–398. doi: 10.1136/ip-2022-044833. [DOI] [PubMed] [Google Scholar]
- 23.Cheng M.-H., Chang S.-F. Frailty as a risk factor for falls among community dwelling people: evidence from a meta-analysis. J Nur Scholarsh. 2017;49(5):529–536. doi: 10.1111/jnu.12322. [DOI] [PubMed] [Google Scholar]
- 24.Zhou Q., Li Y., Gao Q., Yuan H., Sun L., Xi H., et al. Prevalence of frailty among Chinese community-dwelling older adults: a systematic review and meta-analysis. Int J Public Health. 2023;68 doi: 10.3389/ijph.2023.1605964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Johnell O., Kanis J.A. An estimate of the worldwide prevalence and disability associated with osteoporotic fractures. Osteoporos Int. 2006;17(12):1726–1733. doi: 10.1007/s00198-006-0172-4. [DOI] [PubMed] [Google Scholar]
- 26.Hubbard R.E., Rockwood K. Frailty in older women. Maturitas. 2011;69(3):203–207. doi: 10.1016/j.maturitas.2011.04.006. [DOI] [PubMed] [Google Scholar]
- 27.Stevens J.A., Ballesteros M.F., Mack K.A., Rudd R.A., DeCaro E., Adler G. Gender differences in seeking care for falls in the aged Medicare population. Am J Prev Med. 2012;43(1):59–62. doi: 10.1016/j.amepre.2012.03.008. [DOI] [PubMed] [Google Scholar]
- 28.Anbarasan D., Merchant R.A. Association between chronic pain severity, falls, frailty and perceived health in older adults at risk of falls. Eur J Med Res. 2025;30(1):1047. doi: 10.1186/s40001-025-03304-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Stubbs B., Binnekade T., Eggermont L., Sepehry A.A., Patchay S., Schofield P. Pain and the risk for falls in community-dwelling older adults: systematic review and meta-analysis. Arch Phys Med Rehabil. 2014;95(1):175–187.e9. doi: 10.1016/j.apmr.2013.08.241. [DOI] [PubMed] [Google Scholar]
- 30.Huang H., Ni L., Zhang L., Zhou J., Peng B. Longitudinal association between frailty and pain in three prospective cohorts of older population. J Nutr Health Aging. 2025;29(6) doi: 10.1016/j.jnha.2025.100537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Guo X., Tang P., Zhang L., Li R. Tobacco and alcohol consumption and the risk of frailty and falling: a Mendelian randomisation study. J Epidemiol Community Health. 2023;77(6):349–354. doi: 10.1136/jech-2022-219855. [DOI] [PubMed] [Google Scholar]
- 32.Shaper A.G., Wannamethee G., Walker M. Alcohol and mortality in British men: explaining the U-shaped curve. Lancet. 1988;332(8623):1267–1273. doi: 10.1016/s0140-6736(88)92890-5. [DOI] [PubMed] [Google Scholar]
- 33.Ferrucci L., Levine M.E., Kuo P.L., Simonsick E.M. Time and the metrics of aging. Circ Res. 2018;123(7):740–744. doi: 10.1161/CIRCRESAHA.118.312816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.López-Otín C., Blasco M.A., Partridge L., Serrano M., Kroemer G. Hallmarks of aging: an expanding universe. Cell. 2023;186(2):243–278. doi: 10.1016/j.cell.2022.11.001. [DOI] [PubMed] [Google Scholar]
- 35.Tan L.T.P., Lim C.Y.J., Koh J.H., Chandran G., Tan L.F., Merchant R.A. Functional recovery in older adults following mild traumatic brain injury: a systematic review. Eur Geriatr Med. 2026;17(2):507–523. doi: 10.1007/s41999-025-01326-5. [DOI] [PubMed] [Google Scholar]
- 36.De Roza J.G., Ng D.W.L., Mathew B.K., Jose T., Goh L.J., Wang C., et al. Factors influencing fear of falling in community-dwelling older adults in Singapore: a cross-sectional study. BMC Geriatrics. 2022;22(1):186. doi: 10.1186/s12877-022-02883-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.de Souza L.F., Canever J.B., Moreira B.S., Danielewicz A.L., de Avelar N.C.P. Association between fear of falling and frailty in community-dwelling older adults: a systematic review. Clin Interv Aging. 2022;17:129–140. doi: 10.2147/CIA.S328423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Martínez-Arnau F.M., Prieto-Contreras L., Pérez-Ros P. Factors associated with fear of falling among frail older adults. Geriatr Nurs. 2021;42(5):1035–1041. doi: 10.1016/j.gerinurse.2021.06.007. [DOI] [PubMed] [Google Scholar]
- 39.Baek W., Min A., Ji Y., Park C.G., Kang M. Impact of activity limitations due to fear of falling on changes in frailty in Korean older adults: a longitudinal study. Sci Rep. 2024;14(1) doi: 10.1038/s41598-024-69930-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Hartley P., Forsyth F., O’Halloran A., Kenny R.A., Romero-Ortuno R. Eight-year longitudinal falls trajectories and associations with modifiable risk factors: evidence from The Irish Longitudinal Study on Ageing (TILDA) Age Ageing. 2023;52(3) doi: 10.1093/ageing/afad037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Ooi T.C., Rivan N.F.M., Shahar S., Singh D.K.A. Dynamic transitions in falls and physical activity among older adults in Malaysia: a longitudinal multi-state analysis. BMC Public Health. 2025;25(1) doi: 10.1186/s12889-025-25080-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Schoeneberg B., McNeal B., Reisner J., Friesen A., Reed T., Goodwin J.F. The efficacy of perturbation-based balance training among older adults: a systematic review. Phys Occup Ther Geriatr. 2024;42(4):339–355. [Google Scholar]
- 43.Andreoli A., Bianchi A., Campbell A., Bernhardt J., Bayley M., Guo M. In defence of falling: the onomastics and ethics of “therapeutic” falls in rehabilitation. Disabil Rehabil. 2023;45(22):3783–3787. doi: 10.1080/09638288.2022.2135777. [DOI] [PubMed] [Google Scholar]
- 44.Soh S.L.H., McCrum C., Okubo Y., Farlie M., Soh S.E., Tan M.P., et al. About falls efficacy: a commentary on “World guidelines for falls prevention and management for older adults: a global initiative”. J Frailty Sarcopenia Falls. 2024;9(4):281–285. doi: 10.22540/JFSF-09-281. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.