Skip to main content
Download PDF

Social isolation in relation to the incidence and dynamic progression of frailty in the oldest old: a trajectory analysis of a nationwide cohort

BMC Public Health volume 25, Article number: 1809 (2025) Cite this article

Abstract

Background

There is still a lack of evidence examining the association of behavioral and social factors with frailty transitions and mortality. We investigated whether social isolation is associated with different progressions and outcomes of frailty among community-dwelling older adults.

Methods

This community-based cohort study assessed the frailty index and objective social isolation of 31,168 participants (58.3% female; average age: 88.1 ± 11.1 years) from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) from 1998 to 2018. Four Markov state-transition models were constructed to examine the associations between social isolation and the seven transitions of the frailty trajectory.

Results

According to the Markov state-transition model, for every one-point increase in the social isolation score of non-frail participants, the risk of developing prefrailty increased by 4.2% [hazard ratio (HR) = 1.042, 95% confidence interval (CI): 1.007–1.079], whereas for prefrail participants, the risk of developing frailty and death increased by 3.9% (HR = 1.039, 95% CI: 1.007–1.073) and 16.1% (HR = 1.161, 95% CI: 1.099–1.226), respectively. For each increase in the social isolation score in the frail population, the risk of death increased by 2.9% (HR = 1.029, 95% CI: 1.004–1.054). Socially isolated persons had a greater cumulative transition probability to prefrailty and frailty. Socially isolated women were more likely to experience prefrailty and frailty than socially isolated men, whereas the latter were more likely to die from prefrailty and frailty than the former.

Conclusions

This study indicates that social isolation may contribute to an increased risk of both the incidence and progression of frailty, elevating deterioration risks in initially non-frail and prefrail populations, while primarily exacerbating mortality risks in those already experiencing prefrailty or frailty.

Peer Review reports

Background

The oldest-old individuals (≥ 80 years) are the most vulnerable group to intrinsic or extrinsic stressors [1, 2]. Frailty, which is being widely associated with an increased risk of adverse health outcomes [3,4,5,6], is becoming a significant global public health challenge as the global older adult population grows rapidly. However, frailty is not a steady state but rather a dynamic process where improvement is possible, particularly in its early stages [7,8,9,10,11,12]. Although frailty appears to be reversible, worsening frailty is a common frailty transition, while complete remission is a rare event in the absence of an intervention [13, 14]. Identifying and targeting the prognostic factors that might accelerate or mitigate the progression from non-frailty, prefrailty, frailty, and death could aid specific clinical and personal decision-making regarding the timing and types of interventions to ensure effectiveness in managing frail older adults.

Numerous factors, including biological, psychological [15, 16], comorbidity, lifestyle, economic, and social factors [17], have been reported to be significantly associated with frailty [18,19,20]. Accordingly, various long-term modifiable risk factors for frailty, such as overweight/obesity [21, 22], physical inactivity [23, 24], cardiovascular risk [25, 26], alcohol use [27], and environmental influences [28, 29] have been identified. Social isolation, defined as the objective deficiency in social contact and interpersonal interactions, has been robustly linked to increased risks of morbidity and mortality [30]. While the protective effects of social relationships on physical and mental health are well-established [31], the role of social isolation in frailty trajectories and transitions between different stages of frailty remains unclear. Catharine Gale et al. reported that social isolation and loneliness are not associated with a change in the frailty index over a mean period of six years [32]. In contrast, Davies et al. highlighted an association between social isolation and frailty in a cohort of more than 9,000 participants [33]. Several existing longitudinal studies [19, 33,34,35] have examined frailty in primarily Western older adults via the FI [36, 37]. Nonetheless, at present, the evidence concerning the dynamic relationship between social isolation and frailty remains scarce. Accurate information regarding the trajectory from robust to frail to death (the prognosis) is fundamental for appropriate risk stratification [38] and service planning in vulnerable older populations. In addition, previous longitudinal studies [19, 34, 35] have examined frailty transitions primarily in Western older adults; therefore, comparatively little is known about frailty transitions in the older Chinese population. There is also a paucity of literature on frailty progression defined by a cumulative deficit model among community-dwelling older people [14]. Taken together, the association between frailty and social isolation is not fully understood in older adults, especially older adults in China.

Using a nationwide survey over a 20-year follow-up period with the largest dataset of oldest-old cohorts in the world, this study bridges the understanding between social isolation and frailty [1] and defines the relationships among social isolation, frailty progression, and death in advanced age so that relevant assessments and interventions can be targeted appropriately. We hypothesized that different early frailty transition patterns would be associated with social isolation.

Methods

Study design and participants

This longitudinal study used data from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) [39], a large population-based cohort study conducted from 1998 to 2018 by Peking University using face-to-face interviews and self-completed questionnaires [40]. Data from the 1998–2018, 2000–2018, 2002–2018, 2005–2018, and 2008–2018 waves were used. (Fig. S1). We excluded participants who were < 65 or > 106 years old (n = 770), had missing values at baseline for any of the seven social isolation items (n = 1999), had < 80% indicators at baseline to calculate the frailty index (n = 2104), and were lost to follow-up in the subsequent follow-up survey (n = 7350). Overall, 31,168 participants who were followed up for seven waves were included in the analysis (Fig. S1). This study was approved by the Biomedical Ethics Committee of Peking University (IRB00001052–13074). All participants provided written informed consent.

Multidimensional frailty index

The frailty index were constructed following a standard procedure [41]. Multidimensional frailty was measured at baseline and follow-up on the basis of the Rockwood frailty index (FI) (deficit accumulation index), which has been validated in the CLHLS [42,43,44,45]. Our index was composed of 38 deficits, including self-rated health status, interviewer-rated health conditions, comorbidities, cognitive functioning based on Mini-mental state examination, basic activities of daily living measured by the Barthel Activities of Daily Living Test [46], body function, psychological status, sensory function, heart rhythm, and oral health (Table S1). All variables were scored from 0 (absence of each deficit) to 1 (presence of each deficit), while 0.5 indicated an intermediate status. Similarly, variables that were scored on four- or five-point Likert scales were assigned corresponding ordinal values (0, 0.33, 0.67, and 1 on a four-point Likert scale and 0, 0.25, 0.5, 0.75, and 1 on a five-point Likert scale), with larger values indicating more severe impairment (Table S1) [42]. Subsequently, frailty was subsequently categorized using defined cut-off points to indicate individuals who were non-frail (≤ 0.08), prefrail (> 0.08 to < 0.25), or frail (≥ 0.25) [47].

Social isolation score

Social isolation was assessed at baseline according to previously published methods [48, 49], and included four domains: social engagement, living situation, social network, and social support (Table S2; Method S1). The four domains were incorporated into the model as unified social isolation scores due to their weak correlations, which were evaluated using Spearman’s correlation coefficients (Fig. S2). The seven items were summed arithmetically by assigning one point for each item to provide a social isolation score ranging from 0 to 7, with higher scores indicating greater isolation. Using the overall mean social isolation score (mean = 2.76) as a reference, this study defined participants with a social isolation score > 3 as socially isolated.

Mortality

Mortality was measured on the basis of survival status and duration of exposure to death. Survival status was measured on the basis of whether a respondent interviewed in the baseline waves passed away or survived in subsequent waves. The exposure duration for a survivor was measured as the number of months between the interview date at baseline and subsequent waves. For those who passed before a certain wave, the exposure time was measured as the time interval between the date of death and the interview date in the previous wave. The date of death was collected from officially issued death certificates whenever available (> 80% of cases); otherwise the next-of-kin and local residential committees were consulted [50].

Covariates

In the analysis, important covariates with established associations to both frailty and social isolation were considered (Table S3) [51,52,53], including age, sex, residential area, education, household income, childhood socioeconomic status, intake of vegetables and fruits, smoking, drinking, and self-reported regular exercise at baseline. Comorbidities were not included as covariates because the entire set of components was included in the construction of the FI.

Statistical analysis

This study mainly used SPSS (version 26.0) to establish a database, and statistical analysis was conducted using R 4.4.2. The baseline characteristics of the analytical sample were summarized across three frailty states (i.e., non-frailty, prefrailty, and frailty) as percentages for categorical variables, and means and standard deviations for continuous variables. Missing values for the covariates were addressed through multiple imputation.

Seven transition phases were considered based on the natural history of frailty and possible reversal (Fig. 1A). Since frailty is a continuous process, we limited the direct conversion between non-frailty and frailty without prefrailty. This study revealed 48 direct transitions from frailty to non-frailty and 312 deteriorations from non-frailty to frailty (Table S4), which contradicted the frailty transition model we constructed. However, since the cohort was followed up every 3–4 years, prefrailty could occur in the middle of the observation period; thus, 48 transitions to both transitions (from prefrailty to non-frailty or from frailty to prefrailty), and 312 transitions to the two transitions (from non-frailty to prefrailty or from prefrailty to frailty) were incorporated into the analysis. Then, multinomial logit analysis was used to examine the association between social isolation and frailty at baseline.

Fig. 1
figure 1

State transition model and alluvial diagram. A shows the Markov model, with transitions from non-frailty to prefrailty, frailty, and all-cause death. Specific states are reported in boxes, and the transition-specific number of events and percentages (within brackets) are reported on the arrows. B shows the frailty state transitions depicted through the alluvial diagram

Full size image

Markov state-transition models were constructed to examine the associations between social isolation and the seven transitions of the frailty trajectory. Model 1 was unadjusted. Model 2 was adjusted for age at baseline and sex; Model 3 included the same adjustments as Model 2 plus education, residence, and household income; and Model 4 included the same adjustments as Model 3 plus childhood socioeconomic status, regular physical exercise, smoking, drinking, and vegetable and fruit intake. Furthermore, because only a 10-year follow-up period was used for the population enrolled in 2008–2018 (Fig. 1B), cumulative transition probabilities from prior to later states during the 10-year follow-up were predicted for participants with and without social isolation. Differences in cumulative transition probabilities between the sexes were further examined.

Sensitivity analysis

To assess the robustness of the results, several sensitivity analyses were conducted: (1) excluding participants without complete covariate data, the general characteristics of the included and excluded populations are shown in Table S5; (2) redefining the criteria for judging frailty as no frailty (FI ≤ 0.10), prefrailty (FI > 0.10 and < 0.21), and frailty (FI ≥ 0.21) [54]; (3) incorporating education as a continuous variable into the model to counter partial information loss; and (4) investigating the correlation between baseline social isolation status and frailty progression using mixed-effect Poisson regression models.

Results

Descriptive results

A total of 31,168 participants were included in the study, including 4,831 from 1998, 6,224 from 2000, 8,031 from 2002, and 7,067 from 2005, with an average age of 88.1 (SD, 11.1) years at baseline. The average follow-up duration was 4.57 years (SD, 3.92) (last visit -baseline) [142,437.76 person years (PYs)]. At baseline, 5, 138 individuals [aged 89.9 (9.8) years] had social isolation, and 26, 030 had no social isolation [87.7 (11.3) years]. (Table 1). In subsequent years, 4,235 persons transitioned from non-frailty to prefrailty, 5,489 from prefrailty to frailty, 2,227 from prefrailty to non-frailty, 1,502 from frailty to prefrailty, 961 from non-frailty to death, 14,278 from prefrailty to death, and 8,273 from frailty to death (Fig. 1; Table S4). Figure 2 shows the frailty transitions among the population classified by social isolation. Table S6 presents the baseline characteristics of the participants based on frailty status. We also described the time-dependent distribution of social isolation among the 31,168 participants (Fig. S3). The prevalence of social isolation between 1998 and 2018 ranged from 11.9 to 16.8%, respectively. The participants with higher social isolation scores at baseline had greater risks of prefrailty (Relative risk ratio [RRR] = 1.137, 95% CI: 1.091–1.184) and frailty (RRR = 1.265, 95% CI: 1.199–1.335). (Table S7)

Table 1 Characteristics of 31,168 participants by social isolation
Full size table
Fig. 2
figure 2

Transition frequencies of different states in participants by social isolation

Full size image

Associations between the social isolation score and progressions of frailty status

The associations between social isolation and the risk of frailty progression are presented in Table 2. A series of covariates, including age, sex, education level, place of residence, household income, childhood socioeconomic status, exercise, smoking, drinking, and vegetable and fruit intake, were included in the analysis. After adjusting for these potential covariates, for every one-point increase in the social isolation score in the non-frailty population, the risk of developing prefrailty increased by 4.2% (HR = 1.042, 95% CI: 1.007–1.079), whereas for every one-point increase in social isolation in the prefrailty population, the risk of developing frailty and death increased by 3.9% (HR = 1.039, 95% CI: 1.007–1.073) and 16.1% (HR = 1.161, 95% CI: 1.099–1.226), respectively. For every one-point increase in social isolation in the frail population, the risk of death increased by 2.9% (HR = 1.029, 95% CI: 1.004–1.054). However, no statistically significant association was observed between social isolation and improvement in frailty status.

Table 2 Associations between social isolation and risk of seven progressions of frailty using the Markov state-transition model
Full size table

Cumulative transition probability of frailty status

Figure 3 shows the cumulative transition probabilities of different progressions for the participants during the first 10 years. Figure 3A shows the probabilities for all participants of transitioning from non-frailty to prefrailty. The probability for the socially isolated participants was greater than that (0.1-1.1%) for the non-socially isolated participants until the 8.86-year follow-up. The probabilities of transitioning from prefrailty to frailty in socially isolated individuals were greater than those (0.3-1.5%) in non-socially isolated individuals until the 4.62-year follow-up (Fig. 3B). Throughout the follow-up period, the cumulative transition probabilities of transitioning from prefrailty to death (0.9-5.9%) (Fig. 3C) and transitioning from frailty to death (0.8-3.4%) (Fig. 3D) among socially isolated individuals were consistently greater than those in non-isolated individuals.

Fig. 3
figure 3

Cumulative transition probabilities of frailty for participants exposed to the social isolation. Owing to the high number of deaths in the second 10 years of follow-up, this graph shows the cumulative transfer probabilities of frailty states exposed to social isolation (purple) or not (yellow) during the first 10 years, and the differences in cumulative transfer probabilities in females (continuous) and males (dotted). A (for all participants) and 3E (stratified by sex) show the transitions from non-frailty to prefrailty, whereas B (for all participants) and 3 F (stratified by sex) show the transitions from prefrailty to frailty; C (for all participants) and 3G (stratified by sex) show the transitions from prefrailty to death; and D (for all participants) and 3 H (stratified by sex) show the transitions from frailty to death. These models were adjusted for age, sex, education, household income, residence, childhood socioeconomic status, intake of vegetables and fruits, smoking, drinking, and regular physical exercise

Full size image

Compared with men with social isolation, women with social isolation were more likely to experience prefrailty (1.2-5.3%) (Fig. 3E) and frailty (0.2-2.2%) (Fig. 3F). However, men in social isolation were more likely to die from prefrailty (1.4-8.0%) (Fig. 3G) and frailty (3.3-8.6%) (Fig. 3H) than women in social isolation. (Table S8)

Sensitivity analyses

The results of the sensitivity analyses are presented in the Supplementary Materials. First, after excluding participants with missing values (n = 953), the correlation between social isolation and the four transitions remained significant (Model 1 in Table S9). Second, after using another cut-off points to define the different frailty categories, the results remained largely consistent with those of the main models (Model 2 in Table S9). When education level was subsequently considered as a continuous variable in the Markov model, for every one-point increase in social isolation in the prefrail population, the risk of developing frailty and death increased by 5.1% (HR = 1.051, 95% CI: 1.015–1.088) and 12.5% (HR = 1.125, 95% CI: 1.082–1.169), respectively. For every one-point increase in social isolation in the frail population, the risk of death increased by 3.5% (HR = 1.035, 95% CI: 1.006–1.065) (Model 3 in Table S9). Finally, we modeled the associations between social isolation and frailty states over time using mixed-effect Poisson regression models (Table S10) and reported that individuals with social isolation had elevated risks of transitioning from non-frailty to prefrailty (Estimate = 1.016, 95%CI: 1.006–1.027, P = 0.002) and frailty (Estimate = 1.089, 95%CI: 1.066–1.112, P < 0.001), as well as from prefrailty to frailty (Estimate = 1.051, 95%CI: 1.030–1.073, P < 0.001).

Discussion

To the best of our knowledge, this is the first study to use a large sample size and a broad timeframe of 20 years to analyze the associations between social isolation and the incidence and dynamic progression of frailty, as well as the differences in the effects of social isolation on the cumulative transfer probability among different sexes. Social isolation was independently associated with frailty at baseline and at different transition stages from non-frailty to prefrailty, to frailty, and then to death. Participants exposed to social isolation have an increased risk of adverse frailty outcomes. Further sensitivity analyses indicated that even if individuals with social isolation had lower FI at baseline, FI growth rates were faster than those of individuals without social isolation during the follow-up years. Taken together, our results suggest that social isolation, quantified by living situation, social networks, social activities, and social support, can adversely affect frailty transitions.

Notably, starting from the fourth year of follow-up, the number of deaths or losses to follow-up significantly increased; people with higher social isolation scores had lower cumulative transition probabilities than those with non-social isolation because many participants who had higher levels of isolation passed away. In addition, we found that older women were more likely to experience a deterioration of frailty, whereas older men were more likely to die when they had the same status of social isolation. This finding offers the unique insight that although women seem to be more vulnerable in the face of social isolation, men face particularly severe consequences in terms of mortality risk. However, the underlying mechanism remains unclear. A previous study revealed that an inflammatory response is an important pathological mechanism of frailty syndrome [55]. This finding is significant because another study revealed that social isolation is related to the inflammatory response in males and females, whereas acute stressors have a sexually dimorphic effect, enhancing the inflammatory response in females and delaying it in males [56]. Frailty is more prevalent in women (55% vs. 45%) [57], yet men deteriorate faster [58] likely due to hormonal differences—estrogen loss accelerates frailty in women, while testosterone decline exacerbates muscle loss in men [59]. Male frailty is linked to reduced physical activity and cardiovascular risk, whereas female frailty is associated with inflammation and depression [60]. This suggests that the pathophysiological mechanisms are different, Neuroimaging reveals early white matter decline in men and limbic dysfunction in women. Behaviorally, Men benefit more from high-intensity exercise, while women respond better to low-intensity activities [60]. However, men experience earlier physical activity decline [58]. Socially, men’s help lower-seeking behavior increases frailty risk [61], while women’s caregiving roles often compromise their health [62]. Women living alone face a 2.3-fold higher frailty risk than men, reflecting gendered disparities in resource access [63].

Our results are consistent with previous studies reporting that social isolation and frailty are independent factors of mortality [18, 64, 65], and with more theoretical papers or reviews discussing the role that social factors play in predicting the risk of frailty in older adults [66]. In general, prospective longitudinal research with a long follow-up period to assess the effects of social isolation on frailty transition is scarce. Our study identified the relationship between physical isolation and frailty transitions. Compared with similar studies based on the same cohort, this study observed more waves and focused on evaluating the role of social isolation in the seven transitions between the four states. Liu et al. observed only one wave, some transitions in shorter time might not be captured in this study using a 3-year time period [67]. The initiators of this cohort study only observed only the all-cause mortality rate, prevalence of frailty, and disability in older people aged 80 years and above in two 10-year birth cohorts [1].

Different measures of social isolation and frailty can create differences in the observed associations. Cross-study comparisons are limited by inconsistent measurement approaches for both social isolation and frailty. For example, our analysis used a multi-item scoring system adapted to the CLHLS dataset, while other studies may define social isolation using simpler criteria (e.g., monthly in-person contact frequency). Similarly, we assessed frailty through a multidimensional index (FI) covering physical, psychological, and social domains, contrasting with studies relying solely on physical metrics (e.g., gait speed, grip strength) [2]. FI tends to identify a higher prevalence of frailty and capture graded changes over time [14], but these methodological differences create inherent comparability challenges. Additionally, variations in cohort age distributions and measurement protocols further complicate direct comparisons [14]. Interpretation of our findings should account for these measurement-specific contexts.

This study has several major strengths. First, it examines a large population-based cohort across multiple waves and contains detailed information on social isolation (95.4% completion rate), frailty (95.2% completion rate), socioeconomic, lifestyle, and health profile data with the CLHLS. This enabled us to construct a multistate model of frailty development and control for potential confounding factors. Second, unlike traditional Cox regression models, we used the Markov multistate model to distinguish the effects of social isolation on the seven transition phases. Thus, we were able to detect the sensitive stages of frailty development and how its progression could be affected by social isolation. The probability of transition between states, which helps us understand the interrelationship between social isolation and frailty as well as the possible causal relationships between them. Finally, a series of sensitivity analyses confirmed the robustness of the results.

There are also some limitations to this study, such as the following: (1) Some information regarding the evaluations was based on self-reports, which could lead to measurement errors. (2) Some participants were excluded because of missing data or loss to follow-up, which may have led to selection bias. (3) The proportion of socially isolated individuals may have been underestimated because individuals with social isolation or poor health conditions were unlikely to participate in the study. (4) Temporal variations in social isolation during follow-up were not accounted for, but we examined the associations between baseline social isolation and subsequent frailty development and mortality. Most importantly, comorbidity and disabilities are also components of FI. To resolve residual confusion in future research, more detailed health assessments, including comprehensive medical history and diagnostic evaluations, are needed. Future studies should investigate the time-varying nature of social isolation and its dynamic relationships with both incident frailty progression and mortality outcomes.

Conclusion

Social isolation is longitudinally associated with frailty and is a predictor of frailty risk. These findings contribute to the limited evidence regarding the association between social isolation and frailty among community-dwelling older adults. Interventions targeting social isolation hold potential for preventing frailty.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author upon reasonable request.

Abbreviations

HR|:

Hazard ratio

RRR:

Relative risk ratio

CI:

Confidence interval

FI:

Frailty index

SD:

Standard deviation

PYs:

Person-years

SI:

Social isolation

SE:

Standard error

NSI:

Non-social isolation

References

  1. Zeng Y, Feng Q, Hesketh T, Christensen K, Vaupel J. Improvements in survival and activities of daily living despite declines in physical and cognitive functioning among the oldest-old in China – Evidence from a cohort study. Lancet. 2017;389:1619–29.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Bevilacqua G, Jameson KA, Zhang J, Bloom I, Fuggle NR, Patel HP, et al. Relationships between non-communicable disease, social isolation and frailty in community dwelling adults in later life: findings from the Hertfordshire cohort study. Aging Clin Exp Res. 2022;34:105–12.

    Article  PubMed  Google Scholar 

  3. Hoogendijk EO, Afilalo J, Ensrud KE, Kowal P, Onder G, Fried LP. Frailty: implications for clinical practice and public health. Lancet. 2019;394:1365–75.

    Article  PubMed  Google Scholar 

  4. Dent E, Morley JE, Cruz-Jentoft AJ, Woodhouse L, Rodríguez-Mañas L, Fried LP, et al. Physical frailty: ICFSR international clinical practice guidelines for identification and management. J Nutr Health Aging. 2019;23:771–87.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Xue Q-L, Bandeen-Roche K, Tian J, Kasper JD, Fried LP. Progression of physical frailty and the risk of All-Cause mortality: is there a point of no return?? J Am Geriatr Soc. 2021;69:908–15.

    Article  PubMed  Google Scholar 

  6. Solfrizzi V, Scafato E, Custodero C, Piazzolla G, Capogna L, Procaccio A, et al. Biopsychosocial frailty and mild cognitive impairment subtypes: findings from the Italian project on the epidemiology of Alzheimer’s disease (IPREA). Alzheimer’s Dement. 2023;19:3306–15.

    Article  CAS  Google Scholar 

  7. Rockwood K, Mitnitski A. Limits to deficit accumulation in elderly people. Mech Ageing Dev. 2006;127:494–6.

    Article  PubMed  Google Scholar 

  8. Rodríguez Mañas L, García-Sánchez I, Hendry A, Bernabei R, Roller-Wirnsberger R, Gabrovec B, et al. Key messages for a frailty prevention and management policy in Europe from the advantage joint action consortium. J Nutr Health Aging. 2018;22:892–7.

    Article  PubMed  Google Scholar 

  9. Doi T, Makizako H, Tsutsumimoto K, Nakakubo S, Kim M-J, Kurita S, et al. Transitional status and modifiable risk of frailty in Japanese older adults: A prospective cohort study. Geriatr Gerontol Int. 2018;18:1562–6.

    Article  PubMed  Google Scholar 

  10. Kojima G, Taniguchi Y, Iliffe S, Jivraj S, Walters K. Transitions between frailty States among community-dwelling older people: A systematic review and meta-analysis. Ageing Res Rev. 2019;50:81–8.

    Article  PubMed  Google Scholar 

  11. Thompson MQ, Theou O, Adams RJ, Tucker GR, Visvanathan R. Frailty state transitions and associated factors in South Australian older adults. Geriatr Gerontol Int. 2018;18:1549–55.

    Article  PubMed  Google Scholar 

  12. Xue Q-L, Buta B, Ma L, Ge M, Carlson M. Integrating frailty and cognitive phenotypes: why, how, now what?? Curr Geriatr Rep. 2019;8:97–106.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Xue Q-L. The frailty syndrome: definition and natural history. Clin Geriatr Med. 2011;27:1–15.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Kaskirbayeva D, West R, Jaafari H, King N, Howdon D, Shuweihdi F, et al. Progression of frailty as measured by a cumulative deficit index: A systematic review. Ageing Res Rev. 2023;84:101789.

    Article  PubMed  Google Scholar 

  15. Ottenbacher KJ, Graham JE, Al Snih S, Raji M, Samper-Ternent R, Ostir GV, et al. Mexican Americans and frailty: findings from the Hispanic established populations epidemiologic studies of the elderly. Am J Public Health. 2009;99:673–9.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Andrew MK, Fisk JD, Rockwood K. Psychological well-being in relation to frailty: a frailty identity crisis? Int Psychogeriatr. 2012;24:1347–53.

    Article  PubMed  Google Scholar 

  17. Haapanen MJ, Perälä MM, Salonen MK, Kajantie E, Simonen M, Pohjolainen P, et al. Early life determinants of frailty in old age: the Helsinki birth cohort study. Age Ageing. 2018;47:569–75.

    Article  CAS  PubMed  Google Scholar 

  18. Amieva H, Ouvrard-Brouillou C, Dartigues J-F, Pérès K, Tabue Teguo M, Avila-Funes A. Social vulnerability predicts frailty: towards a distinction between fragility and frailty?? J Frailty Aging. 2022;11:318–23.

    Article  CAS  PubMed  Google Scholar 

  19. Pollack LR, Harrison SL, Cawthon PM, Ensrud K, Lane NE, Barrett-Connor E, et al. Patterns and predictors of frailty transitions in older men: the osteoporotic fractures in men study. J Am Geriatr Soc. 2017;65:2473–9.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Pek K, Chew J, Lim JP, Yew S, Tan CN, Yeo A, et al. Social frailty is independently associated with mood, nutrition, physical performance, and physical activity: insights from a Theory-Guided approach. Int J Environ Res Public Health. 2020;17:4239.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Strandberg TE, Stenholm S, Strandberg AY, Salomaa VV, Pitkälä KH, Tilvis RS. The obesity paradox, frailty, disability, and mortality in older men: A prospective, longitudinal cohort study. Am J Epidemiol. 2013;178:1452–60.

    Article  PubMed  Google Scholar 

  22. Stenholm S, Guralnik JM, Bandinelli S, Ferrucci L. The prognostic value of repeated measures of lower extremity performance: should we measure more than once?? J Gerontol Biol Sci Med Sci. 2014;69:894–9.

    Article  Google Scholar 

  23. Savela SL, Koistinen P, Stenholm S, Tilvis RS, Strandberg AY, Pitkälä KH, et al. Leisure-Time physical activity in midlife is related to old age frailty. Journals Gerontology: Ser A. 2013;68:1433–8.

    Google Scholar 

  24. Abe T, Nofuji Y, Seino S, Murayama H, Yoshida Y, Tanigaki T, et al. Healthy lifestyle behaviors and transitions in frailty status among independent community-dwelling older adults: the Yabu cohort study. Maturitas. 2020;136:54–9.

    Article  PubMed  Google Scholar 

  25. Strandberg TE, Sirola J, Pitkälä KH, Tilvis RS, Strandberg AY, Stenholm S. Association of midlife obesity and cardiovascular risk with old age frailty: a 26-year follow-up of initially healthy men. Int J Obes. 2012;36:1153–7.

    Article  CAS  Google Scholar 

  26. Atkins JL, Delgado J, Pilling LC, Bowman K, Masoli JAH, Kuchel GA, et al. Impact of low cardiovascular risk profiles on geriatric outcomes: evidence from 421,000 participants in two cohorts. J Gerontol Biol Sci Med Sci. 2019;74:350–7.

    Article  Google Scholar 

  27. Strandberg AY, Trygg T, Pitkälä KH, Strandberg TE. Alcohol consumption in midlife and old age and risk of frailty: alcohol paradox in a 30-year follow-up study. Age Ageing. 2018;47:248–54.

    Article  PubMed  Google Scholar 

  28. Hall A, Boulton E, Kunonga P, Spiers G, Beyer F, Bower P, et al. Identifying older adults with frailty approaching end-of-life: A systematic review. Palliat Med. 2021;35:1832–43.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Di Ciaula A, Portincasa P. The environment as a determinant of successful aging or frailty. Mech Ageing Dev. 2020;188:111244.

    Article  PubMed  Google Scholar 

  30. Hoang P, King JA, Moore S, Moore K, Reich K, Sidhu H, et al. Interventions associated with reduced loneliness and social isolation in older adults. JAMA Netw Open. 2022;5:e2236676.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Fried L, Prohaska T, Burholt V, Burns A, Golden J, Hawkley L, et al. A unified approach to loneliness. Lancet. 2020;395:114.

    Article  PubMed  Google Scholar 

  32. Gale CR, Westbury L, Cooper C. Social isolation and loneliness as risk factors for the progression of frailty: the english longitudinal study of ageing. Age Ageing. 2018;47:392–7.

    Article  PubMed  Google Scholar 

  33. Davies K, Maharani A, Chandola T, Todd C, Pendleton N. The longitudinal relationship between loneliness, social isolation, and frailty in older adults in England: a prospective analysis. Lancet Healthy Longev. 2021;2:e70–7.

    Article  PubMed  Google Scholar 

  34. Gill TM, Gahbauer EA, Allore HG, Han L. Transitions between frailty States among Community-Living older persons. Arch Intern Med. 2006;166:418–23.

    Article  PubMed  Google Scholar 

  35. Fallah N, Mitnitski A, Searle SD, Gahbauer EA, Gill TM, Rockwood K. Transitions in frailty status in older adults in relation to mobility: A Multi-State modeling approach employing a deficit count. J Am Geriatr Soc. 2011;59:524–9.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Kojima G, Taniguchi Y, Aoyama R, Tanabe M. Associations between loneliness and physical frailty in community-dwelling older adults: A systematic review and meta-analysis. Ageing Res Rev. 2022;81:101705.

    Article  PubMed  Google Scholar 

  37. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: evidence for a phenotype. Journals Gerontology: Ser A. 2001;56:M146–57.

    CAS  Google Scholar 

  38. Collins GS, Reitsma JB, Altman DG, Moons KGM. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. BMJ. 2015;350:g7594.

    Article  PubMed  Google Scholar 

  39. Zeng Y. Healthy longevity in China: demographic, socioeconomic, and psychological dimensions. Dordrecht? Springer; 2008.

  40. Lv Y, Yang Z, Ye L, Jiang M, Zhou J, Guo Y, et al. Long-term fine particular exposure and incidence of frailty in older adults: findings from the Chinese longitudinal healthy longevity survey. Age Ageing. 2023;52:afad009.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Searle SD, Mitnitski A, Gahbauer EA, Gill TM, Rockwood K. A standard procedure for creating a frailty index. BMC Geriatr. 2008;8:24.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Yang F, Gu D. Predictability of frailty index and its components on mortality in older adults in China. BMC Geriatr. 2016;16:145.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Dupre ME, Gu D, Warner DF, Yi Z. Frailty and type of death among older adults in China: prospective cohort study. BMJ. 2009;338:b1175.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Zeng Y, Feng Q, Gu D, Vaupel J. Demographics, phenotypic health characteristics and genetic analysis of centenarians in China. Mech Ageing Dev. 2017;165 Pt B:86–97.

    Article  Google Scholar 

  45. Gu Y, Wu W, Bai J, Chen X, Chen X, Yu L, et al. Association between the number of teeth and frailty among Chinese older adults: a nationwide cross-sectional study. BMJ Open. 2019;9:e029929.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Mahoney FI, Barthel DW. FUNCTIONAL EVALUATION: THE BARTHEL INDEX. Md State Med J. 1965;14:61–5.

    CAS  PubMed  Google Scholar 

  47. Romero-Ortuno R. An alternative method for frailty index cut-off points to define frailty categories. Eur Geriatr Med. 2013;4.

  48. Kelly ME, Duff H, Kelly S, McHugh Power JE, Brennan S, Lawlor BA, et al. The impact of social activities, social networks, social support and social relationships on the cognitive functioning of healthy older adults: a systematic review. Syst Rev. 2017;6:259.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Yu B, Steptoe A, Chen Y. Social isolation, loneliness, and all-cause mortality: A cohort study of 35,254 Chinese older adults. J Am Geriatr Soc. 2022;70:1717–25.

    Article  PubMed  Google Scholar 

  50. Gu D, Feng Q. Frailty still matters to health and survival in centenarians: the case of China. BMC Geriatr. 2015;15:159.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Wennberg AM, Ebeling M, Ek S, Meyer A, Ding M, Talbäck M, et al. Trends in frailty between 1990 and 2020 in Sweden among 75-, 85-, and 95-Year-Old women and men: A nationwide study from Sweden. J Gerontol Biol Sci Med Sci. 2022;78:342–8.

    Article  Google Scholar 

  52. Cao W, Cao C, Zheng X, Ji K, Liang Q, Wu Y, et al. Factors associated with medication adherence among Community-Dwelling older people with frailty and Pre-Frailty in China. Int J Environ Res Public Health. 2022;19:16001.

    Article  PubMed  PubMed Central  Google Scholar 

  53. Dugravot A, Fayosse A, Dumurgier J, Bouillon K, Rayana TB, Schnitzler A, et al. Social inequalities in Multimorbidity, frailty, disability, and transitions to mortality: a 24-year follow-up of the Whitehall II cohort study. Lancet Public Health. 2019;5:e42–50.

    Article  PubMed  PubMed Central  Google Scholar 

  54. Rockwood K, Song X, Mitnitski A. Changes in relative fitness and frailty across the adult lifespan: evidence from the Canadian National population health survey. CMAJ. 2011;183:E487–494.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Walston J, McBurnie MA, Newman A, Tracy RP, Kop WJ, Hirsch CH, et al. Frailty and activation of the inflammation and coagulation systems with and without clinical comorbidities: results from the cardiovascular health study. Arch Intern Med. 2002;162:2333–41.

    Article  PubMed  Google Scholar 

  56. Hermes GL, Rosenthal L, Montag A, McClintock MK. Social isolation and the inflammatory response: sex differences in the enduring effects of a prior stressor. Am J Physiology-Regulatory Integr Comp Physiol. 2006;290:R273–82.

    Article  CAS  Google Scholar 

  57. Lu Y-W, Chang C-C, Chou R-H, Lee W-J, Chen L-K, Huang P-H, et al. Sex differences in the frailty phenotype and mortality in the I-Lan longitudinal aging study cohort. BMC Geriatr. 2024;24:182.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Isernia S, Cazzoli M, Baglio G, Cabinio M, Rossetto F, Giunco F, et al. Differential roles of neural integrity, physical activity and depression in frailty: Sex-Related differences. Brain Sci. 2023;13:950.

    Article  PubMed  PubMed Central  Google Scholar 

  59. Tazzeo C, Rizzuto D, Calderón-Larrañaga A, Dekhtyar S, Zucchelli A, Xia X, et al. Living longer but frailer?? Temporal trends in life expectancy and frailty in older Swedish adults. J Gerontol Biol Sci Med Sci. 2024;79:glad212.

    Article  Google Scholar 

  60. Reid N, Young A, Shafiee Hanjani L, Hubbard RE, Gordon EH. Sex-specific interventions to prevent and manage frailty. Maturitas. 2022;164:23–30.

    Article  PubMed  Google Scholar 

  61. Li Y, Du Z, Kondo N. Effect modification of social participation in the relationship between living arrangements and frailty among older adults in Japan: differences based on gender. Arch Gerontol Geriatr. 2024;117:105231.

    Article  PubMed  Google Scholar 

  62. Julià M, Méndez-Rivero F, Gómez-Gómez Á, Pozo ÓJ, Bolíbar M. Association between precarious employment and chronic stress: effect of gender, stress measurement and precariousness Dimensions-A Cross-Sectional study. Int J Environ Res Public Health. 2022;19:9099.

    Article  PubMed  PubMed Central  Google Scholar 

  63. Mufina Begam J, Abdul Azeez EP, Irshad CV. Padma Sri Lekha P. Social frailty moderates the association between gender and life satisfaction: evidence from a Large-Scale survey among Indian older adults. J Gerontol Soc Work. 2025;:1–20.

  64. Andrew MK, Mitnitski AB, Rockwood K. Social vulnerability, frailty and mortality in elderly people. PLoS ONE. 2008;3:e2232.

    Article  PubMed  PubMed Central  Google Scholar 

  65. Ouvrard C, Avila-Funes JA, Dartigues J-F, Amieva H, Tabue-Teguo M. The social vulnerability index: assessing replicability in predicting mortality over 27 years. J Am Geriatr Soc. 2019;67:1305–6.

    Article  PubMed  Google Scholar 

  66. Freer K, Wallington SL. Social frailty: the importance of social and environmental factors in predicting frailty in older adults. Br J Community Nurs. 2019;24:486–92.

    Article  PubMed  Google Scholar 

  67. Liu Z, Wei Y, Wei L, Jiang X, Wang X, Shi Y, et al. Frailty transitions and types of death in Chinese older adults: a population-based cohort study. Clin Interv Aging. 2018;13:947–56.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We gratefully thank the CLHLS research team at the Peking University of Healthy Aging and Development (http://chads.nsd.pku.edu.cn) for providing the publicly available dataset.

Funding

This study was partly supported by the National Natural Science Foundation of China (82071272, 82171491, and 72174159).

Author information

Author notes

  1. Junyi Wang and Yafu Yu Joint first authors.

  2. Yan Zeng and Wei Tan contributed equally to this work.

Authors and Affiliations

  1. Geriatric Hospital Affiliated to Wuhan University of Science and Technology, West Huangjiahu Road, Hongshan District, Wuhan, 430065, China

    Junyi Wang, Yafu Yu, Lang Xu, Xinyan Xie, Xiaochang Liu, Juan Zhou, Guirong Cheng, Feifei Hu, Dan Liu, Linya Huang, Gangbin Han, Shiyue Li, Dan Song, Jing Liu, Qianqian Nie, Cheng Cai, Yuyang Cui, Wei Tan & Yan Zeng

  2. Brain Science and Advanced Technology Institute, Wuhan University of Science and Technology, West Huangjiahu Road, Hongshan District, Wuhan, 430065, China

    Junyi Wang, Yafu Yu, Lang Xu, Xinyan Xie, Xiaochang Liu, Juan Zhou, Guirong Cheng, Feifei Hu, Dan Liu, Linya Huang, Gangbin Han, Shiyue Li, Dan Song, Jing Liu, Qianqian Nie, Cheng Cai, Yuyang Cui, Wei Tan & Yan Zeng

Authors
  1. Junyi Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Yafu Yu
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Lang Xu
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Xinyan Xie
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Xiaochang Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Juan Zhou
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Guirong Cheng
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Feifei Hu
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Dan Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Linya Huang
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Gangbin Han
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. Shiyue Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  13. Dan Song
    View author publications

    You can also search for this author inPubMed Google Scholar

  14. Jing Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  15. Qianqian Nie
    View author publications

    You can also search for this author inPubMed Google Scholar

  16. Cheng Cai
    View author publications

    You can also search for this author inPubMed Google Scholar

  17. Yuyang Cui
    View author publications

    You can also search for this author inPubMed Google Scholar

  18. Wei Tan
    View author publications

    You can also search for this author inPubMed Google Scholar

  19. Yan Zeng
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

YZ, JYW, and YFY designed the study. XCL, JZ, LYH, GBH, SYL, DS, JL, QQN, CC, and YYC contributed to the literature search. YZ, WT, LX, XYX, FFH, DL, and GRC supervised the study. JYW and YFY performed the sample acquisition and data collection; JYW and YFY performed the data analysis; YZ, JYW, YFY, LX and XYX interpreted the results of the analysis. YZ, JYW and YFY wrote the original manuscript; YZ, WT, LX, XYX, FFH, DL, and GRC reviewed and edited the manuscript. YZ and WT acquired funding and administered the project. All the authors have read and contributed to the manuscript.

Corresponding authors

Correspondence to Wei Tan or Yan Zeng.

Ethics declarations

Ethics approval and consent to participate

The biomedical ethics committee of Peking University approved the study (IRB00001052-13074), and all participants or their legal guardians provided written informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wang, J., Yu, Y., Xu, L. et al. Social isolation in relation to the incidence and dynamic progression of frailty in the oldest old: a trajectory analysis of a nationwide cohort. BMC Public Health 25, 1809 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12889-025-22596-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12889-025-22596-5

Keywords

BMC Public Health

ISSN: 1471-2458

Contact us