Highlights
-
•
Higher ACEs scores were associated with an increased probability of forward transition (worsening) in frailty states.
-
•
Higher ACEs scores were associated with a decreased probability of backward transition of frailty states.
-
•
Social participation was associated with an increased probability of backward transition.
-
•
Social participation may help older adults with ACEs exposure mitigate worsening frailty development.
Keywords: Frailty, ACEs, Social participation, Multi-state model
Abstract
Objectives
Adverse childhood experiences (ACEs) are associated with frailty, while the association with frailty state transitions and the role of social participation remain unclear. This study aimed to investigate the association between ACEs and frailty state transitions, alongside the moderating effect of social participation
Methods
Data from 9,621 adults aged 45 and older from the China Health and Retirement Longitudinal Study (2011–2020) were analyzed. Frailty was measured with the frailty index, while ACEs and social participation were measured with a validated questionnaire. The association between ACEs and frailty state transitions was estimated using multi-state models. An interaction analysis were used to examine the moderating effects of social participation.
Results
Participants with higher ACEs scores (≥4) were associated with an increased probability of forward transition (robust to pre-frail, HR = 1.37, 95%CI: 1.21–1.54; prefrail to frail, HR = 1.39, 95%CI: 1.18–1.63) and decreased probability of backward transition (pre-frail to robust, HR = 0.64, 95%CI: 0.55–0.76). Additionally, participants with moderate and high level social participation were associated with an increased probability of backward transition (pre-frail to robust, HR = 1.11, 95%CI: 1.01–1.23; frail to pre-frail, HR = 1.17, 95%CI: 1.02–1.33, respectively). Social participation moderated the association between ACEs exposure and frailty (P for interaction <0.05), while participants with lower ACEs scores (1 and 2) and high social participation were associated with an increased probability of transition from frail to pre-frail (HR = 1.26, 95%CI: 1.04–1.89 and HR = 1.15, 95%CI: 1.08–1.69).
Conclusions
High ACEs scores were associated with an increased likelihood of adverse frailty development. Older adults with ACEs exposure might benefit from intervention strategies to improve social participation.
1. Introduction
Frailty is characterized by increased vulnerability due to age-related decline in multiple physiological systems, and it is associated with adverse health outcomes such as falls, delirium, and disability [1]. Given the aging global population, frailty poses a significant challenge to health systems worldwide. A meta-analysis conducted across 62 countries and territories found that the prevalence of frailty was 24% among individuals aged 50 years and older [2]. In China, the prevalence of frailty and pre-frailty among community-dwelling older adults aged 65 and older is 10% and 43%, respectively [3]. Additionally, frailty is associated with multi-domain risk factors, including demographic, health-related, and physical factors [4], yet its etiology remains unclear.
Adverse childhood experiences (ACEs) refer to traumatic events during childhood linked to various adverse health outcomes later in life, including chronic diseases, mental health disorders, and early mortality [5,6]. ACEs are associated with adverse outcomes across multiple physiological systems, potentially linking them to frailty. A population-based study found that individuals exposed to ACEs had elevated levels of frailty, measured using frailty index (FI) [7]. Additionally, a prospective cohort study found that ACEs were associated with an increased incidence of frailty among the oldest old [8]. In China, cumulative ACEs were associated with a greater number of frailty events and a faster decline trajectory in FI among middle-aged and older adults [9].
In addition to the adverse health outcomes resulting from ACEs, studies have also examined moderating factors such as social participation, which involves engaging in activities with social interaction [10]. Social participation is associated with various health benefits [11] and can mitigate the negative impacts of ACEs on depression [12] and heavy alcohol consumption behavior [13]. Research shows that exercise-based social participation and high self-rated health are associated with reversing frailty progression [14]. Additionally, longitudinal studies have found that social participation was inversely associated with frailty risk among middle-aged and older adults [15,16].
Although evidence has established a relationship between ACEs and the incidence and frailty [8,9], it remains unclear whether ACEs are associated with the transition between frailty states. The dynamic nature of frailty development, which involves both forward or backward transitions [1,17], highlights the importance of identifying potential factors involved during this process. Our study aims to address these gaps by utilizing large cohort data of middle-aged and older adults in China. We hypothesize that exposure to ACEs increases the probability of worsening frailty states and that social participation moderates this association. The findings of this study will provide essential insights into the mechanisms underlying the association between ACEs and frailty in middle-aged and older adults, potentially informing interventions aimed at preventing or mitigating potential negative health outcomes later in life.
2. Method
2.1. Sample and design
We used data from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative longitudinal survey of participants aged 45 and older and their spouses. The first wave of the survey began in 2011, including 17,708 participants from 150 counties across 28 provinces. Follow-up surveys were conducted in 2013, 2015, 2018, and 2020. Besides, the CHARLS life history survey was conducted in 2014 to collect information on early life experiences through a retrospective design. Detailed information on the survey design has been published previously [18].
For this study, we used data from CHARLS wave 1 (2011) to wave 5 (2020), as the frailty measurements were obtained from these waves. The ACEs information was collected from the 2014 life history survey. Of the 17,708 participants from the baseline survey, we excluded those who were under 45 in 2011. Additionally, participants with missing information on ACEs and frailty or those were lost to follow-up during waves 1–5 were excluded. This resulted in a final sample of 9,621 participants. Detailed information about the sample selection is presented in Fig. 1.
Fig. 1.
Diagram of participants selection process.
2.2. Frailty states
We measured frailty states using the frailty index, which represents a proportion of health deficits accumulated by an individual across multiple domains. The frailty index provides a comprehensive geriatric assessment of frailty and is useful for continuous follow-up, sensitive to modification, and adaptable to different populations [19]. In this study, we calculated the frailty index based on the cumulative health deficits of each participant [20], and the construction of the frailty index followed a standard procedure proposed by previous research [21].
Consistent with previous research [19,22,23] and considering the availability of data, we included 29 health deficit items from 6 domains at each wave to calculate the frailty index. The six domains of health deficits include diseases, self-rated health, activities of daily living, instrumental activities of daily living, cognition, and psychosocial well-being. While previous studies have recommended including at least 30 items in the FI for comparability and reliability [24], when considering depression and cognition as composite measures, as done in prior studies, the total number of deficits in our frailty index increases to 41.Detailed information on the description and manipulation of each item is presented in e-Table S1 in the supplementary file. We dichotomized each deficit item into 0–1 binary variables, with 1 indicating a deficit. Furthermore, we calculated the frailty index by dividing the total deficit counts by 29 for each participant. Finally, we categorized the frailty index into three levels of states, and the cut-off points were followed by previous research [23]. Specifically, we defined three frailty states: robust (frailty index ≤0.10), prefrail (frailty index >0.10 to <0.25), and frail (frailty index ≥0.25), respectively.
2.3. Adverse childhood experiences
The construction of the ACEs variable has been described in our previous publication based on CHARLS [25]. In accordance with the previous study [5,26], we conceptualized 13 ACEs items from 4 domains, including abuse and neglect (physical abuse and emotional neglect), household dysfunction (domestic violence, household substance abuse, household mental illness, incarcerated household member, and parental separation or divorce), family member death or illness (parental death, sibling death, parental disability), and adverse peer or neighborhood environment (bullying, peer rejection, unsafe neighborhood). We measured the ACEs by first dichotomizing each item into 0–1 binary variables, then counted the total items (a higher count denotes severe ACEs exposure). We then created a categorical variable to indicate the ACEs scores. To address potential data sparsity at higher ACE scores and to align with previous research indicating a threshold effect at four or more ACEs, we created a categorical variable indicating ACE scores of 0, 1, 2, 3, and ≥4 [5,27].
2.4. Social participation
Social participation in middle-aged and older adults can be conceptualized as conscious and active engagement in outdoor interacting activities with others, which can bring personal satisfaction [10]. Therefore, based on the previous study [28], we measured social participation by self-reported participation in the following six activities, including interacted with friends; played Ma-jong, played chess, played cards, or went to community club; went to a sport, social, or other kind of club; took part in a community-related organization; done voluntary or charity; and attended an educational or training course. To measure the degree of social participation, we considered both the type and frequency of each activity, with frequency was coded as follows: 0 = None, 1 = Not Regularly, 2 = Almost Every Week, and 3 = Almost Daily. Following the approach of a previous study [29], we calculated social participation index was by summing the frequency scores across all six activities for each participant. Subsequently, participants were categorized into three groups—low, moderate, and high social participation—based on tertiles of the social participation index. Moreover, social participation was considered time-variant and measured across wave 1 to wave 5.
2.5. Covariates
We included demographic characteristics, lifestyle, and wealth as covariates based on prior studies. For demographic characteristics, we included age, gender, marital status, region, and education level. As lifestyle variables, we included current smoking and drinking status (yes/no). In addition, we used the annual per capita household expenditure level as the proxy variable for wealth. All these covariates were collected from the CHARLS 2011 baseline wave and treated as categorical variables.
2.6. Statistical analysis
We performed a descriptive analysis to describe the characteristics of our samples by frailty status at baseline, with all the categorical data were presented as numbers and proportions. Additionally, we used the univariate Chi-square test to assess the differences between each characteristic and frailty status. To visualize the longitudinal transition patterns of frailty states, we created an alluvial chart using the R ggalluvial package [30].
We used the multi-state Markov model to estimate the changes in frailty states, as our data followed a panel structure with participants interviewed at arbitrary continuous times. First, we estimated the transition intensity and transition probability during follow-up time intervals between states (the allowed transitions were: Robust → Pre-Frail, Pre-Frail → Robust, Pre-Frail → Frail, and Frail → Pre-Frail). Second, we estimated the association between ACEs and the transitions of frailty states after adjusting for covariates. The multi-state model was fitted using R msm package [31], and the results were presented as hazard ratios (HR) and 95% confidence intervals (95% CI).
Previous literature suggests that social participation might mitigate the negative effects of ACEs on health. Therefore, we further performed an interaction analysis on social participation and tested the interaction between ACEs and social participation using the likelihood ratio test.
Regarding the missing values, we conducted multivariate imputation by chained equations [32] using the random forest method [33] with missing values in frailty items and covariates. All statistical analyses were conducted using R Statistical Software (4.4.1) [34], and P < 0.05 was considered statistically significant.
3. Results
3.1. Baseline characteristics
Table 1 shows the baseline characteristics of participants, stratified by frailty states. The study included 9,621 participants, among whom 4,683 were classified as robust, 4,089 as pre-frail, and 849 as frail. There were significant differences in the baseline characteristics of the participants across the three frailty states. Specifically, the participants with frailty had higher ACEs scores, were older, female, had lower education levels, lived in the rural areas, and had lower annual per capita household expenditure levels. In addition, participants who smoked, drank and lacked social participation also tended to be frail.
Table 1.
Baseline characteristics of participants, by frailty states.
| Variables | Robust (n = 4,683) | Pre-frail (n = 4,089) | Frail (n = 849) | All (n = 9,621) | P-value |
|---|---|---|---|---|---|
| ACEs score | <0.001 | ||||
| 0 | 560 (11.96) | 323 (7.90) | 52 (6.12) | 935 (9.72) | |
| 1 | 1,106 (23.62) | 753 (18.42) | 127 (14.96) | 1,986 (20.64) | |
| 2 | 1,223 (26.12) | 974 (23.82) | 207 (24.38) | 2,404 (24.99) | |
| 3 | 873 (18.64) | 851 (20.81) | 182 (21.44) | 1,906 (19.81) | |
| 4- | 921 (19.67) | 1,188 (29.05) | 281 (33.10) | 2,390 (24.84) | |
| Age (years) | <0.001 | ||||
| 45- | 1,273 (27.18) | 732 (17.90) | 76 (8.95) | 2,081 (21.63) | |
| 50- | 870 (18.58) | 632 (15.46) | 105 (12.37) | 1,607 (16.70) | |
| 55- | 1,082 (23.10) | 969 (23.70) | 178 (20.97) | 2,229 (23.17) | |
| 60- | 739 (15.78) | 801 (19.59) | 203 (23.91) | 1,743 (18.12) | |
| 65- | 392 (8.37) | 517 (12.64) | 129 (15.19) | 1,038 (10.79) | |
| 70- | 217 (4.63) | 260 (6.36) | 95 (11.19) | 572 (5.95) | |
| 75- | 110 (2.35) | 178 (4.35) | 63 (7.42) | 351 (3.65) | |
| Sex | <0.001 | ||||
| Male | 2,179 (46.53) | 2,308 (56.44) | 566 (66.67) | 5,053 (52.52) | |
| Female | 2,504 (53.47) | 1,781 (43.56) | 283 (33.33) | 4,568 (47.48) | |
| Marital status | <0.001 | ||||
| Current married | 630 (13.45) | 662 (16.19) | 169 (19.91) | 1,461 (15.19) | |
| Others | 4,053 (86.55) | 3,427 (83.81) | 680 (80.09) | 8,160 (84.81) | |
| Education level | <0.001 | ||||
| Illiterate | 992 (21.18) | 1,143 (27.95) | 346 (40.75) | 2,481 (25.79) | |
| Can read | 798 (17.04) | 849 (20.76) | 200 (23.56) | 1,847 (19.20) | |
| Elementary school | 1,039 (22.19) | 945 (23.11) | 164 (19.32) | 2,148 (22.33) | |
| Middle school | 1,203 (25.69) | 771 (18.86) | 99 (11.66) | 2,073 (21.55) | |
| High school and more | 651 (13.90) | 381 (9.32) | 40 (4.71) | 1,072 (11.14) | |
| Region | <0.001 | ||||
| Rural | 1,700 (36.30) | 1,341 (32.80) | 220 (25.91) | 3,261 (33.89) | |
| Urban | 2,983 (63.70) | 2,748 (67.20) | 629 (74.09) | 6,360 (66.11) | |
| Annual per capita household expenditure level | <0.001 | ||||
| Q1 | 1,121 (23.94) | 1,041 (25.46) | 253 (29.80) | 2,415 (25.10) | |
| Q2 | 1,115 (23.81) | 1,071 (26.19) | 210 (24.73) | 2,396 (24.90) | |
| Q3 | 1,209 (25.82) | 988 (24.16) | 209 (24.62) | 2,406 (25.01) | |
| Q4 | 1,238 (26.44) | 989 (24.19) | 177 (20.85) | 2,404 (24.99) | |
| Smoke now | <0.001 | ||||
| No | 3,029 (64.68) | 2,933 (71.73) | 671 (79.03) | 6,633 (68.94) | |
| Yes | 1,654 (35.32) | 1,156 (28.27) | 178 (20.97) | 2,988 (31.06) | |
| Drink now | <0.001 | ||||
| No | 2,862 (61.11) | 2,838 (69.41) | 686 (80.80) | 6,386 (66.38) | |
| Yes | 1,821 (38.89) | 1,251 (30.59) | 163 (19.20) | 3,235 (33.62) | |
| Social participation | <0.001 | ||||
| Low | 1,492 (31.85) | 1,394 (34.10) | 321 (37.81) | 3207 (33.33) | |
| Moderate | 1,553 (33.16) | 1,360 (33.27) | 294 (34.63) | 3207 (33.33) | |
| High | 1,638 (34.98) | 1,335 (32.63) | 234 (27.56) | 3207 (33.33) |
Notes: Data are presented as n (%). The chi-square test was used for categorical variables. ACEs variables came from the 2014 survey, and higher score indicated a higher adversity level. The annual per capita household expenditure level was categorized by quantile of the annual per capita household expenditure (Q1 = quantile 1 [the lowest] to Q4 = quantile4 [the highest]).
Abbreviation: ACEs = Adverse childhood experiences.
3.2. The transition of frailty status during follow-up
Fig. 2 shows the longitudinal transition patterns of frailty states during a ten-year follow-up. e-Table 2 shows that participants in robust and pre-frail states often remained in the same state from one assessment to another, with a transition probability of 74% for remaining robust or remaining pre-frail. Regarding forward transitions, the transition probability for robust to pre-frail and pre-frail to frail states was 24% and 14%. Regarding backward transitions, the transition probability for frail to pre-frail and pre-frail to robust states was 25% and 11%. Of note, direct transitions from robust to frail or frail to robust were rare.
Fig. 2.
Alluvial chart of the longitudinal transitions of frailty states in CHARLS (n = 9,621).
Note: The height of the stacked bars each year is the number of participants identified as belonging to this state. The thickness of the streams connecting the stacked bars between years is the number of participants transitioning from the states.
3.3. ACEs, social participation and transition of frailty states
Table 2 shows the results of the multi-state model after adjusting for covariates. Compared with those without ACEs exposure, participants with ACEs scores of 3 and ≥4 were associated with an increased probability of transition from robust to pre-frail (HR = 1.20, 95%CI: 1.06–1.36 and HR = 1.37, 95%CI: 1.21–1.54, respectively). Similarly, compared with those without ACEs exposure, participants with ACEs scores of 3 and ≥4 were associated with an increased probability of transition from pre-frail to frail (HR = 1.21, 95%CI: 1.02–1.43 and HR = 1.39, 95%CI: 1.18–1.63, respectively). In contrast, participants with ACEs scores of 3, and ≥4 were associated with a decreased probability of transition from pre-frail to robust (HR = 0.78, 95%CI: 0.66 – 0.93 and HR = 0.64, 95%CI: 0.55–0.76, respectively) compared with those without ACEs exposure.
Table 2.
Hazard ratios and 95% confidence intervals for the association between adverse childhood experiences and frailty state transition over ten years.
| Variables | Robust->prefrail | Prefrail->frail | Prefrail->robust | Frail->prefrail |
|---|---|---|---|---|
| ACEs score | HR (95%CI) | HR (95%CI) | HR (95%CI) | HR (95%CI) |
| 0 | Ref. | Ref. | Ref. | Ref. |
| 1 | 1.04 (0.92,1.17) | 1.00 (0.84,1.19) | 0.94 (0.80,1.10) | 1.05 (0.82,1.34) |
| 2 | 1.10 (0.98,1.24) | 1.05 (0.89,1.25) | 0.85 (0.72,1.00) | 1.09 (0.86,1.37) |
| 3 | 1.20 (1.06,1.36) | 1.21 (1.02,1.43) | 0.78 (0.66,0.93) | 0.87 (0.69,1.11) |
| 4- | 1.37 (1.21,1.54) | 1.39 (1.18,1.63) | 0.64 (0.55,0.76) | 0.83 (0.66,1.04) |
| Social participation | ||||
| Low | Ref. | Ref. | Ref. | Ref. |
| Moderate | 1.01 (0.93−1.09) | 0.99 (0.90−1.09) | 1.11 (1.01−1.23) | 1.07 (0.94−1.21) |
| High | 0.94 (0.90−1.05) | 0.99 (0.90−1.09) | 0.94 (0.84−1.04) | 1.17 (1.02−1.33) |
Note: HR and (95%CI) are presented; the Model was adjusted for age, sex, marital status, region, education, smoking, drink, and annual per capita household expenditure level.
Abbreviation: ACEs, Adverse childhood experiences; HR, Hazard Ratio; CI, Confidence Interval.
Furthermore, compared with those with low social participation, participants with moderate social participation were associated with an increased probability of transition from pre-frail to robust (HR = 1.11, 95%CI: 1.01–1.23). Additionally, participants with high social participation were associated with an increased probability of transition from frail to pre-frail (HR = 1.17, 95% CI: 1.02–1.33).
3.4. ACEs × Social participation interaction and transition of frailty states
Based on the above results, we further examined the interaction between ACEs and social participation. Table 3 shows the results of the interaction analysis, and there was a statistically significant interaction between ACEs and social participation in the adjusted models.
Table 3.
Hazard ratios and 95% confidence intervals for the interaction between ACEs score and social participation on transitions of frailty states over ten years.
| Variables |
Robust->prefrail | Prefrail->frail | Prefrail->robust | Frail->prefrail | |
|---|---|---|---|---|---|
| Social participation | ACEs score | HR (95%CI) | HR (95%CI) | HR (95%CI) | HR (95%CI) |
| Low | 0 | Ref. | Ref. | Ref. | Ref. |
| Low | 1 | 1.02 (0.83−1.26) | 1.06 (0.78−1.44) | 0.95 (0.71−1.27) | 0.89 (0.60−1.33) |
| Low | 2 | 1.06 (0.86−1.30) | 1.10 (0.82−1.48) | 0.85 (0.64−1.13) | 0.99 (0.68−1.46) |
| Low | 3 | 1.14 (0.92−1.41) | 1.18 (0.88−1.58) | 0.76 (0.57−1.02) | 0.73 (0.50−1.08) |
| Low | 4- | 1.38 (1.12−1.70) | 1.35 (1.02−1.79) | 0.64 (0.48−0.85) | 0.79 (0.55−1.15) |
| Moderate | 0 | 0.96 (0.75−1.23) | 0.80 (0.55−1.17) | 1.03 (0.74−1.44) | 1.07 (0.67−1.72) |
| Moderate | 1 | 0.99 (0.80−1.22) | 0.91 (0.66−1.24) | 1.04 (0.78−1.38) | 0.95 (0.63−1.42) |
| Moderate | 2 | 1.07 (0.87−1.31) | 1.08 (0.80−1.45) | 0.99 (0.75−1.31) | 1.02 (0.70−1.50) |
| Moderate | 3 | 1.27 (1.02−1.57) | 1.24 (0.92−1.66) | 0.85 (0.64−1.14) | 0.87 (0.59−1.29) |
| Moderate | 4- | 1.36 (1.10−1.68) | 1.45 (1.09−1.92) | 0.68 (0.51−0.92) | 0.80 (0.55−1.17) |
| High | 0 | 0.93 (0.73−1.17) | 1.15 (0.82−1.61) | 0.99 (0.71−1.37) | 0.81 (0.48−1.36) |
| High | 1 | 0.98 (0.79−1.21) | 1.03 (0.75−1.41) | 0.85 (0.63−1.14) | 1.26 (1.04−1.89) |
| High | 2 | 1.06 (0.86−1.30) | 0.98 (0.73−1.32) | 0.74 (0.55−1.00) | 1.15 (1.08−1.69) |
| High | 3 | 1.07 (0.86−1.33) | 1.20 (0.89−1.63) | 0.77 (0.57−1.03) | 0.97 (0.65−1.44) |
| High | 4- | 1.21 (0.98−1.50) | 1.36 (1.02−1.81) | 0.63 (0.47−0.85) | 0.81 (0.56−1.19) |
Note: Likelihood ratio test was used for the interaction test, and the P for interaction <0.05.
HR and (95%CI) are presented; Model 1 was adjusted for age, sex, marital status, region, education, smoking, drink, and annual per capita household expenditure level.
Abbreviation: ACEs = Adverse childhood experiences; HR= Hazard Ratio; CI = Confidence Interval.
Compared with participants with low social participation and without ACEs exposure, ACEs scores of ≥4 were associated with an increased probability of transition from robust to pre-frail (HR = 1.38, 95%CI: 1.12–1.70) and transition from pre-frail to frail (HR = 1.35, 95%CI: 1.02–1.79). Besides, among participants with moderate social participation, participants with ACEs scores of 3 and ≥4 were associated with an increased probability of transition from robust to pre-frail (HR = 1.27, 95%CI: 1.02–1.57 and HR = 1.36, 95%CI: 1.10–1.68, respectively). Additionally, ACEs scores of ≥4 were also associated with an increased probability transition from pre-frail to frail (HR = 1.45, 95%CI: 1.09–1.92). Furthermore, among participants with high social participation, ACEs score of ≥4 were associated with an increased probability transition from pre-frail to frail (HR = 1.36, 95%CI: 1.02–1.81).
In contrast, compared with participants with low social participation and without ACEs exposure, ACEs scores of ≥4 among the low, moderate and high social participation groups were associated with a decreased probability of transition from pre-frail to robust (HR = 0.64, 95%CI: 0.48 – 0.85;HR = 0.68, 95%CI: 0.51 – 0.92 and HR = 0.63, 95%CI: 0.47−0.85, respectively).
Moreover, participants with high social participation and lower ACEs scores (1 and 2) were associated with an increased probability of transition from frail to pre-frail (HR = 1.26, 95%CI: 1.04–1.89 and HR = 1.15, 95%CI: 1.08–1.69, respectively).
4. Discussion
4.1. Main findings
In this study, we used a large cohort of middle-aged and older adults from 2011 to 2020 and found that participants with higher ACEs scores (3 and ≥4) were associated with an increased probability of forward transition(worsening) and a decreased probability of backward transition(improvement) of frailty states. Additionally, participants with lower ACEs scores were more likely to transition from frail to pre-frail, particularly among those with high social participation. These results suggest that social participation might moderate the relationship between ACEs scores and frailty transitions.
Previous studies have shown that ACEs are associated with frailty, including an elevated frailty index, increased frailty risk, and worse frailty trajectory [[7], [8], [9]]. This study extends the current literature by demonstrating an association between ACEs and the transition between frailty states. Participants with higher ACEs scores were more likely to experience worsening transitions and less likely to experience improving transitions. The association between ACEs and late-life frailty can be explained by the biological embedding theory, which suggests that ACEs might result in long-term physiological dysregulation across multiple systems, including the immune, neural, and metabolic systems [35]. This physiological mechanism could lead to the adverse development of frailty [36]. ACEs have also been linked to health risk behaviors in adulthood [37], which could further increase the risk of frailty. Additionally, adulthood adversity might play a critical role in the association between ACEs and frailty [38]. ACEs are associated with lower education, unemployment, and heavy drinking [39], suggesting that both ACEs and adulthood adversities can shape an individual's capacity to attain healthy lifestyles and outcomes.
Regarding the role of social participation, previous studies have found that it is associated with reduced risks of frailty and reversing frailty progression among middle-aged and older adults [[14], [15], [16]]. Our findings extend this literature by showing that individuals with high social participation and lower ACEs scores were more likely to experience a backward transition. Moreover, previous studies have shown that social participation could moderate the negative impacts of ACEs on health outcomes and behavior, including depression and heavy alcohol consumption [12,13]. Our study expands on these findings by demonstrating that participants with lower ACEs scores and engaged in high social participation were more likely to transition from frail to pre-frail. Exposure to ACEs can lead to inadequate emotional and material support and heightened stress, which can result in maladaptive coping strategies such as smoking and heavy drinking. In contrast, social support can provide additional emotional sustenance and active coping assistance [40], thereby moderating the negative effect of ACEs [41].
However, our findings also highlight that the buffering effect of social participation is limited to individuals with lower ACEs exposure. Participants with high ACEs score is associated with a higher probability forward transition and lower probability of transitioning from pre-frail to robust, regardless of social participation levels. This suggests that early-life adversity may have long-lasting consequences to physiological disruptions, which predispose individuals to various health problems throughout the life course [35]. Our finding is also in line with prior research that social support during pregnancy provides a protective effect against the negative effects of maternal ACEs on infant birth size, especially for those with low and moderate ACEs exposure [42].
4.2. Strengths and limitations
Our study has several strengths. First, it utilized nationally representative data and extended the population to middle age, which extends the findings and inferences to a broader population, making it more generalizable. Second, the incorporation of a transition model approach enables the investigation of both improvement and deterioration in health outcomes, making it distinctive from previous studies that focus on one-directional approaches. Finally, our study investigated the role of social participation and its moderating effects on health outcomes, offering insights into effective intervention strategies. This can assist policymakers in designing effective health interventions that increase social participation, improve health outcomes, and reduce health inequalities.
However, we also acknowledge some limitations. The retrospective nature of ACEs data collection may introduce potential recall bias, which could impact the data's accuracy. Nonetheless, previous studies have validated similarly collected adult-reported ACEs information, showing that such bias does not invalidate the measurement [43]. Additionally, longitudinal studies often face issues such as participants loss to follow-up, which may introduce selection bias and potentially affect the external validity of our study. However, we addressed this limitation by employing imputation with random forest methods in the statistical analysis, enhancing the robustness of the imputation model and reducing bias in our findings. Finally, given the observational design of our study, we acknowledge that our findings indicate an association rather than causation. Thus, while our study provides insight into the impact of social participation on health outcomes in middle-aged and older adults, further research with more rigorous designs is needed to establish causation.
5. Conclusion
Our study provides evidence that higher ACEs scores were associated with an increased probability of forward transition and a decreased probability of backward transition of frailty states among middle-aged and older adults. Engagement in social participation acts as a protective factor, mitigating the association between exposure to ACEs and transitions in frailty. These findings highlight the need for developing interventions that target the physiological and psychological processes underlying the association between ACEs and frailty, as well as promoting social participation in middle-aged and older adults as a potential strategy to mitigate adverse health outcomes associated with ACEs exposure. Further research is needed to elucidate the mechanisms underlying the ACEs-frailty pathway and identify effective interventions for preventing or reducing the risk of frailty among middle-aged and older adults.
CRediT authorship contribution statement
Jiajia Li & Heming Pei: literature search, writing (original draft); Jiajia Li, Xiaojin Yan and Yue Wei statistical analysis and visualization, Lijun Pei & Gong Chen: conceptualization, funding acquisition, and writing (review and revised the article critically for important intellectual content).
Ethic approval
The study was approved by the Institutional Review Board at Peking University. The IRB approval number for the main household survey, including anthropometrics, is IRB00001052-11015. Written consent was obtained from all participants or their legal representatives.
Funding sources
This work was supported by National Social Science Fund of China (grant numbers: 23ZDA101), National Natural Science Foundation of China (grant numbers: 41871360), National Key Research and Development Program of China (No. 2018YFC1004303).
Availability of data and materials
The data are available from the websites of China Health and Retirement Longitudinal Study
Declaration of competing interest
None declared.
Acknowledgment
The authors gratefully acknowledge the National School of Development of Peking University for providing the data of China Health and Retirement Longitudinal Study (CHARLS).
Footnotes
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jnha.2024.100400.
Contributor Information
Gong Chen, Email: chengong@pku.edu.cn.
Lijun Pei, Email: peilj@pku.edu.cn.
Appendix A. Supplementary data
The following is Supplementary data to this article:
References
- 1.Clegg A., Young J., Iliffe S., Rikkert M.O., Rockwood K. Frailty in elderly people. Lancet. 2013;381:752–762. doi: 10.1016/S0140-6736(12)62167-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.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. 2021;50:96–104. doi: 10.1093/ageing/afaa219. [DOI] [PubMed] [Google Scholar]
- 3.He B., Ma Y., Wang C., Jiang M., Geng C., Chang X., et al. Prevalence and risk factors for frailty among community-dwelling older people in China: a systematic review and meta-analysis. J Nutr Health Aging. 2019;23:442–450. doi: 10.1007/s12603-019-1179-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Qin Y., Hao X., Lv M., Zhao X., Wu S., Li K. A global perspective on risk factors for frailty in community-dwelling older adults: a systematic review and meta-analysis. Arch Gerontol Geriatr. 2023;105 doi: 10.1016/j.archger.2022.104844. [DOI] [PubMed] [Google Scholar]
- 5.Felitti V.J., Anda R.F., Nordenberg D., Williamson D.F., Spitz A.M., Edwards V., et al. Relationship of Childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. Am J Prev Med. 1998;14:245–258. doi: 10.1016/s0749-3797(98)00017-8. [DOI] [PubMed] [Google Scholar]
- 6.Hughes K., Ford K., Bellis M.A., Glendinning F., Harrison E., Passmore J. Health and financial costs of adverse childhood experiences in 28 European countries: a systematic review and meta-analysis. Lancet Public Health. 2021;6:e848–e857. doi: 10.1016/s2468-2667(21)00232-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mian O., Anderson L.N., Belsky D.W., Gonzalez A., Ma J., Sloboda D.M., et al. Associations of adverse childhood experiences with frailty in older adults: a cross-sectional analysis of data from the Canadian longitudinal study on aging. Gerontology. 2022;68:1091–1100. doi: 10.1159/000520327. [DOI] [PubMed] [Google Scholar]
- 8.Dimitriadis M.M., Jeuring H.W., Marijnissen R.M., Wieringa T.H., Hoogendijk E.O., Oude Voshaar R.C. Adverse Childhood Experiences and frailty in later life: a prospective population-based cohort study. Age Ageing. 2023;52 doi: 10.1093/ageing/afad010. afad010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wang Q. Association of adverse childhood experiences with frailty index level and trajectory in China. JAMA Netw Open. 2022;5 doi: 10.1001/jamanetworkopen.2022.25315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dehi M., Mohammadi F. Social participation of older adults: a concept analysis. Int J Community Based Nurs Midwifery. 2020:8. doi: 10.30476/ijcbnm.2019.82222.1055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wanchai A., Phrompayak D. Social participation types and benefits on health outcomes for elder people: a systematic review. Ageing Int. 2019;44:223–233. doi: 10.1007/s12126-018-9338-6. [DOI] [Google Scholar]
- 12.Nishio M., Green M., Kondo N. Roles of participation in social activities in the association between adverse childhood experiences and health among older Japanese adults. SSM - Popul Health. 2022;17 doi: 10.1016/j.ssmph.2021.101000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ashaba S., Kakuhikire B., Baguma C., Satinsky E.N., Perkins J.M., Rasmussen J.D., et al. Adverse childhood experiences, alcohol consumption, and the modifying role of social participation: population-based study of adults in southwestern Uganda. SSM - Ment Health. 2022;2 doi: 10.1016/j.ssmmh.2022.100062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Takatori K., Matsumoto D. Social factors associated with reversing frailty progression in community-dwelling late-stage elderly people: an observational study. PLoS One. 2021;16 doi: 10.1371/journal.pone.0247296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sun J., Kong X., Li H., Chen J., Yao Q., Li H., et al. Does social participation decrease the risk of frailty? Impacts of diversity in frequency and types of social participation on frailty in middle-aged and older populations. BMC Geriatr. 2022;22:1–12. doi: 10.1186/s12877-022-03219-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ge L., Yap C.W., Heng B.H. Associations of social isolation, social participation, and loneliness with frailty in older adults in Singapore: a panel data analysis. BMC Geriatr. 2022;22:26. doi: 10.1186/s12877-021-02745-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.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;50:81–88. doi: 10.1016/j.arr.2019.01.010. [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;43:61–68. doi: 10.1093/ije/dys203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Cesari M., Gambassi G., Abellan van Kan G., Vellas B. The frailty phenotype and the frailty index: different instruments for different purposes. Age Ageing. 2014;43:10–12. doi: 10.1093/ageing/aft160. [DOI] [PubMed] [Google Scholar]
- 20.Mitnitski A.B., Mogilner A.J., Rockwood K. Accumulation of deficits as a proxy measure of aging. Sci World J. 2001;1:323–336. doi: 10.1100/tsw.2001.58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Searle S.D., Mitnitski A., Gahbauer E.A., Gill T.M., Rockwood K. A standard procedure for creating a frailty index. BMC Geriatr. 2008;8:24. doi: 10.1186/1471-2318-8-24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fan J., Yu C., Guo Y., Bian Z., Sun Z., Yang L., et al. Frailty index and all-cause and cause-specific mortality in Chinese adults: a prospective cohort study. Lancet Public Health. 2020;5:e650–e660. doi: 10.1016/S2468-2667(20)30113-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lu Z., Er Y., Zhan Y., Deng X., Jin Y., Ye P., et al. Association of frailty status with risk of fall among middle-aged and older adults in China: a nationally representative cohort study. J Nutr Health Aging. 2021;25:985–992. doi: 10.1007/s12603-021-1655-x. [DOI] [PubMed] [Google Scholar]
- 24.Nguyen Q.D., Moodie E.M., Keezer M.R., Wolfson C. Clinical correlates and implications of the reliability of the frailty index in the Canadian longitudinal study on aging. J Gerontol Ser A. 2021;76:e340–e346. doi: 10.1093/gerona/glab161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Li J., Lin S., Yan X., Pei L., Wang Z. Adverse childhood experiences and trajectories of ADL disability among middle-aged and older adults in China: findings from the CHARLS cohort study. J Nutr Health Aging. 2022;26:1034–1041. doi: 10.1007/s12603-022-1863-z. [DOI] [PubMed] [Google Scholar]
- 26.Afifi T.O. In: Adverse Childhood Experiences. Asmundson G.J.G., Afifi T.O., editors. Academic Press; 2020. Chapter 3 - Considerations for expanding the definition of ACEs; pp. 35–44. [Google Scholar]
- 27.Hughes K., Bellis M.A., Hardcastle K.A., Sethi D., Butchart A., Mikton C., et al. The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis. Lancet Public Health. 2017;2:e356–e366. doi: 10.1016/s2468-2667(17)30118-4. [DOI] [PubMed] [Google Scholar]
- 28.Yang R., Wang H., Edelman L.S., Tracy E.L., Demiris G., Sward K.A., et al. Loneliness as a mediator of the impact of social isolation on cognitive functioning of Chinese older adults. Age Ageing. 2020;49:599–604. doi: 10.1093/ageing/afaa020. [DOI] [PubMed] [Google Scholar]
- 29.Fang B., Huang J., Zhao X., Liu H., Chen B., Zhang Q. Concurrent and lagged associations of social participation and frailty among older adults. Health Soc Care Community. 2022:30. doi: 10.1111/hsc.13888. [DOI] [PubMed] [Google Scholar]
- 30.Bojanowski M., Edwards R. 2016. Alluvial: R Package for Creating Alluvial Diagrams. [Google Scholar]
- 31.Sherris M., Wei P. A multi-state model of functional disability and health status in the presence of systematic trend and uncertainty. North Am Actuar J. 2021;25:17–39. doi: 10.1080/10920277.2019.1708755. [DOI] [Google Scholar]
- 32.Buuren S.V., Groothuis-Oudshoorn K. Mice: multivariate imputation by chained equations in R. J Stat Softw. 2010:1–68. [Google Scholar]
- 33.Shah A.D., Bartlett J.W., Carpenter J., Nicholas O., Hemingway H. Comparison of random forest and parametric imputation models for imputing missing data using MICE: a CALIBER study. Am J Epidemiol. 2014;179:764–774. doi: 10.1093/aje/kwt312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.R Core Team . R Foundation for Statistical Computing; Vienna, Austria: 2022. R: A Language and Environment for Statistical Computing. [Google Scholar]
- 35.Berens A.E., Jensen S.K.G., Nelson C.A. Biological embedding of childhood adversity: from physiological mechanisms to clinical implications. BMC Med. 2017;15:135. doi: 10.1186/s12916-017-0895-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Fried L.P., Xue Q.-L., Cappola A.R., Ferrucci L., Chaves P., Varadhan R., et al. Nonlinear multisystem physiological dysregulation associated with frailty in older women: implications for etiology and treatment. J Gerontol A Biol Sci Med Sci. 2009;64A:1049–1057. doi: 10.1093/gerona/glp076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Campbell J.A., Walker R.J., Egede L.E. Associations between adverse childhood experiences, high-risk behaviors, and morbidity in adulthood. Am J Prev Med. 2016;50:344–352. doi: 10.1016/j.amepre.2015.07.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Jones T.M., Nurius P., Song C., Fleming C.M. Modeling life course pathways from adverse childhood experiences to adult mental health. Child Abuse Negl. 2018;80:32–40. doi: 10.1016/j.chiabu.2018.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Lin W.-H., Chiao C. The relationship between adverse childhood experience and heavy smoking in emerging adulthood: the role of not in education, employment, or training status. J Adolesc Health. 2022;70:155–162. doi: 10.1016/j.jadohealth.2021.07.022. [DOI] [PubMed] [Google Scholar]
- 40.Thoits P.A. Mechanisms linking social ties and support to physical and mental health. J Health Soc Behav. 2011;52:145–161. doi: 10.1177/0022146510395592. [DOI] [PubMed] [Google Scholar]
- 41.Sheffler J.L., Piazza J.R., Quinn J.M., Sachs-Ericsson N.J., Stanley I.H. Adverse childhood experiences and coping strategies: identifying pathways to resiliency in adulthood. Anxiety Stress Coping. 2019;32:594–609. doi: 10.1080/10615806.2019.1638699. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Appleton A.A., Kiley K., Holdsworth E.A., Schell L.M. Social support during pregnancy modifies the association between maternal adverse childhood experiences and infant birth size. Matern Child Health J. 2019;23:408–415. doi: 10.1007/s10995-018-02706-z. [DOI] [PubMed] [Google Scholar]
- 43.Hardt J., Rutter M. Validity of adult retrospective reports of adverse childhood experiences: review of the evidence. J Child Psychol Psychiatry. 2004;45:260–273. doi: 10.1111/j.1469-7610.2004.00218.x. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data are available from the websites of China Health and Retirement Longitudinal Study