Changes in frailty and depressive symptoms among middle-aged and elderly Chinese: a nationwide cohort study

doi:10.21203/rs.3.rs-3218980/v1

Download PDF

Research Article

Changes in frailty and depressive symptoms among middle-aged and elderly Chinese: a nationwide cohort study

https://doi.org/10.21203/rs.3.rs-3218980/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 25 Jan, 2024

Read the published version in BMC Public Health →

You are reading this latest preprint version

Background and Aims:

The elderly population bears a severe burden of disease due to frailty and depressive symptoms, however, the results of association between the two in the elderly Chinese population have been conflicting. Therefore, this study aimed to investigate the developmental trajectories and interactions of frailty and depressive symptoms in the Chinese middle-aged and elderly population.

Methods

The study used three waves of data from 2011, 2013, and 2015 in the China Health and Retirement Longitudinal Study (CHARLS) database, focused on middle-aged and elderly people ≥ 45 years of age, and analyzed using latent growth models and cross-lagged models.

Results

According to the parallel latent growth model's findings, the initial level of depressive symptoms had a significant positive predictive effect on the initial level of frailty and a significant negative predictive effect on the rate of change in frailty. A substantial positive predictive relationship existed between the rate of change in depressive symptoms and the rate of change in response to frailty. The overall sample population, the middle-aged population, and the elderly population all showed evidence of a bidirectional causal connection between frailty and depressive symptoms as a result of cross-lagged analyses.

Conclusions

This study recommended advancing the age of concern for frailty and depressive symptoms to middle-aged adults. In the middle-aged and elderly Chinese population, the higher the initial level of frailty and the faster the rate of change, the higher the initial level of depressive symptoms and the faster the rate of change.

Frailty

Depressive symptoms

Chinese middle-aged and older adults

CHARLS

China is the most populous country in the world and the country with the largest total elderly population in the world^[1]. According to the National Bureau of Statistics of China, by the end of 2022, the population aged 60 years and older in China was 280 million, accounting for 19.8% of the total population^[2]. China is also aging significantly more rapidly than other countries, having entered a deeply aging society in only 21 years. The problems faced by the aging society are not completely the increased burden of old-age care, or face more severe health problems of the elderly.

Older persons frequently experience frailty and depression, which come with a significant illness burden. Among them, frailty is a clinical state characterized by decreased physical physiological reserve, diminished stress resistance, increased vulnerability of the body and susceptibility to diseases^[3]. Frailty is mostly related to advancing age, and its prevalence is fast increasing as the world's aging population expands^[4]. Frailty also increases the risk of adverse consequences of falls, hospitalization, and death in older adults^[4]. Meanwhile, depression is regarded as the most common psychological problem in older adults, with up to one-third of all older adults experiencing depressive symptoms, yet its symptoms are often frequently simply ignored and inadequately treated^[5], leading to family breakdown, disability, and the result of worsening disease, and increasing mortality^[6].

According to previous studies, there is a high degree of overlap in diagnostic criteria for frailty and depression^[7], such as fatigue, but they also have their own unique symptoms. And therefore, symptom overlap cannot accurately account for the association between these two diseases; there may be some connection between them. Several earlier studies have shown a heightened correlation between debilitation and depression^{[8, 9]}. Frail patients have an increased chance of experiencing depression, as do depressed patients^[8]. In addition, results from a prospective cohort study pointed to frailty as an independent predictor of depressive symptoms in community-dwelling older adults^[10]. Results from another prospective study supported the association of late-life depression with frailty in the following year^[11]. Thus, there may be a causal relationship between frailty and depression. The strong association between frailty and depression can be explained in several ways: First both two share common risk factors, such as functional impairment and poor financial status^[12–14]. Second, the two also share common pathophysiological pathways, such as the inflammatory system^[15]. These studies suggested that frailty and depression may be "overlap syndromes" or that there may be some relationship between the two. In addition, because both frailty and early depressive symptoms were reversible, understanding the relationship between the two was important for early screening and treatment.

Considering the prevalence of frailty and depression in the elderly population and the range of adverse consequences for older adults, it is critical to further explore the relationship between the two. Two previous studies have been conducted in China on the relationship between frailty and depression in the Chinese elderly population, but the findings on the association between the two are somewhat controversial. A bidirectional causal connection between frailty and depressive symptoms was demonstrated in a longitudinal study of older adults aged 70–84 in rural Jiangsu Province, China, using cross-lagged panel models^[16]. The findings of this study may not be generally applicable to the whole older adult population in China because it was limited to a rural population of elderly adults aged 70 to 84 in one area. Another cohort study using the CHARLS database found that frailty was an independent predictor of subsequent depression among older (60 years and above) Chinese adults, but no association between frailty and depressive symptoms was found^[17]. Additionally, the bidirectional causal association between frailty and depressive symptoms could not be investigated by the statistical method used in this study; it could only look at predictors. Consequently, in order to better understand how the two interact dynamically and how this affects the process of personal growth, more potent statistical methods should be used.

Frailty is an age-related geriatric syndrome^[18], and a recent study reported that the prevalence of pre-frailty and frailty in people aged 40–49 years was 45% and 1.4%, respectively, equivalent to those aged 70–75 years^[19] suggesting that we should start preventing frailty at the age of 40. Additionally, a recent study has revealed that middle-aged persons in the UK population have a more significant link between frailty and mental health than older adults^[20]. This showed that the association between frailty and depressive symptoms may be more pronounced in middle-aged people. However, there are currently no study on the association between frailty and depressive symptoms in the Chinese middle-aged population, and there is a dearth of knowledge on the intervention variables that target frailty in middle-aged adults.

Therefore, the purpose of this study was to further examine: (1) the trajectory of debilitating and depressive symptoms in the Chinese middle-aged and elderly population; (2) the causal direction of frailty and depressive symptoms in the Chinese middle-aged and elderly population. This information can serve as a scientific guide for the early use of preventative strategies and therapeutic regimens for incapacitating and depressed symptoms in the Chinese middle-aged and elderly population.

2.1 Data Source and Participants

The China Health and Retirement Longitudinal Study (CHARLS) was the source of all the information used in this study^[21]. CHARLS is a high-quality public database designed to investigate families and individuals of people aged 45 years and older, using multistage PPS sampling, a household survey of middle-aged and elderly population in 28 provincial administrative regions of China covering 450 communities (villages) in 150 counties and districts of 28 provinces across the country, including personal basic information, family structure and economic support, health status and physical measurements^[21]. The baseline survey was conducted in 2011 and tracked biennially, with participants followed up in 2013, 2015, and 2018, and phase 4 data have now been published and generally represent the Chinese elderly population. CHARLS was approved by the Biomedical Ethics Committee of Peking University and informed consent was obtained from all participants.

Data from 2011 (first survey, 17,708 participants), 2013 (second survey), and 2015 (third survey) CHARLS database were used in this study. The following were the inclusion requirements for participants in this study: (1) Age ≥ 45 years; (2) Complete general demographic data, including gender, education, and marital status; (3) Complete information on frailty and depressive symptoms could be provided; (4) Participants participated in all three surveys.

Figure 1 shows the selection process for study participants. A total of 17,708 participants were available at baseline (2011), 17280 participants were ≥ 45 years of age, a final total of 13,042 participants were finally included in the trial at baseline after excluding 4238 people who had missing demographic, depression symptoms, and frailty data at baseline, 11735 participants attended all three surveys after excluding 1307 participants who were lost during follow-up in 2013 or 2015, 2692 participants with missing general data and 2998 participants with missing depressive symptoms, and the study eventually comprised 9281 participants altogether (Fig. 1).

2.2 Measurements

2.2.1 Frailty

The frailty index (FI) was measured to estimate the frailty, which calculated the accumulation of age-related health defects, and its index can be programmed and freely constructed according to needs on the basis of following the health defect selection principle, but it should contain at least 30 health defect items^[22]. In this study, 40 indexes were selected to construct the FI assessment scale according to the criteria for constructing FI^[23]: 6 items in Basic Activity of Daily Living (BADL), 5 entries for Instrumental Activities of Daily Living Scale (IADL), hearing disorder, visual impairment, language disorder, 14 chronic diseases (Hypertension, Dyslipidemia, Diabetes or high blood sugar, Cancer or malignant tumor, Chronic lung diseases, Liver disease, Heart attack, Stroke, Kidney disease, Stomach or other digestive diseases, Emotional, nervous, or psychiatric problems, Memory-related disease, Arthritis or rheumatism, Asthma), 2 disabilities (Physical disabilities and Brain damage/mental retardation), Self-assessed health status and health changes, 3 mobility, 5 muscle performance and cognitive function measured using telephone interview for cognitive status-modified (TICS-m). TICS-m includes daily memory ability, immediate memory ability, delayed memory ability, calculation and drawing ability, with a total of 21 questions, and a total score of 0 to 21 points, with higher scores having a better cognitive function, and health deficit scores taking the value of the actual score/21 (the list of frailty indexes was shown in Supplementary Table 1). The evaluation of the above dimensions assigned health variables 0, 1 (0 = no health deficit, 1 = health deficit) according to the type of variable, etc. FI was calculated by dividing the health deficit score by the total number of participants (40 in this study), which ranged from 0 to 1, with larger values indicating more weakness. FI has already been used in several CHARLS studies^{[24, 25]}.

2.2.2 Depressive symptoms

Depressive symptoms in CHARLS were assessed using the 10-item Center for Epidemiological Studies Depression Scale (CESD-10), which used a Likert score of 0 to 3 to assess the frequency of symptoms experienced by participants over the past 1 week, with the following response options: Rarely or none of the time represented ≤ 1 day and was assigned a value of 0; Some or a little of the time represented 1–2 days and was assigned a score of 1; Occasionally or a moderate amount of the time represents 3–4 days and was assigned a value of 2; Most or all of the time represented 5–7 days and was assigned a value of 3. Among them, questions 5 and 8 were entitled reverse scoring questions, and responses were assigned reverse. The total score on the scale ranged from 0 to 30, with higher scores indicating more severe depressive symptoms. CES-D-10 had been validated in the elderly Chinese population^[26].

2.2.3 Demographic variables

Demographic variables in this study included age; sex (male; female); Marital status (Married with spouse present; Married but not living with spouse temporarily for reasons such as work; Separated; Divorced; Widowed; Never married); Education (No formal education (illiterate); Elementary school or below; Middle school and above); Smoking was assessed based on whether or not smoking had occurred (yes; no); Alcohol consumption was assessed according to whether or not alcohol was consumed in the past year and frequency of alcohol consumption (Drink more than once a month; Drink but less than once a month; None of these).

2.4 Statistical Analysis

Firstly, Means ± Standard Deviations (M ± SD) and frequencies (percentages) were used to describe the basic characteristics of study participants. Then, the correlation between frailty and depressive symptoms was determined using Spearman correlation analysis (frailty and depressive symptoms were tested to be skewed distributions). Then Latent Growth Modeling (LGM) was constructed to explore the trends of frailty and depression symptoms in the middle-aged and elderly population. LGM is an important statistical method for exploring the developmental trajectories of variables and is generally modeled by data from three or more time points^[27]. In LGM, the intercept represents the initial level of the variable and the slope represents the rate of change of the variable^[28]. Finally, a cross-lagged panel model (CLPM) was created to see if there was a causal relationship between frailty and depressive symptoms in the elderly Chinese population. Because females had a predilection for depressive symptoms, we divided the total sample into two groups: male and female and established CLPM separately. CLPM was a longitudinal analytical model that investigated the inter-predictive relationship or quasi-causal relationship between variables^[29]. According to the measurement of two variables at different time points, the estimated path coefficients of the influence of the pretest variables on the posttest variables had a temporal order relationship, which was in line with the principle of epidemiological causal inference. Several correlation coefficients were obtained by calculation. Among them, comparing the absolute value of standardized coefficients could investigate the strength of the causal relationship between variables. Model goodness of fit was assessed using the following indices^[30]: Comparative fit index (CFI), Tucker-Lewis index (TLI), Root mean square error of approximation (RMSEA) and Standardized root mean square residual error (SRMR) were compared for evaluation. Where CFI > 0.9, TLI > 0.9, RMSEA < 0.08, SRMR < 0.08 indicate a good model fit. Considering multiple repeated measurements, P < 0.01 was considered statistically significant. R 4.2.1 was used to accomplish the data cleaning. R software was utilized for the remainder of the statistical study, with the exception of Mplus, which was employed to create the latent growth model. The R packages used for these analyses were lavaan, tidySEM, and corrplot.

3.1 Descriptive Analysis

A total of 9281 participants were included in the study in the total sample, including 5535 middle-aged adults aged 45–59 years and 3746 older adults ≥ 60 years. 49.30% of the participants in the total sample were male, 3734 participants smoked, and 65.50% never drank alcohol. By education level, only 22.60% of participants in the total sample were illiterate. Table 1 shows the basic information for the total sample, the 45–59 age group (middle-aged) and the ≥ 60 age group (elderly) (Table 1).

Table 1

Descriptive statistics for 9281 study participants
Variables		Total	45–59	≥ 60
N		9281	5535	3746
Age (M ± SD)		57.97 ± 8.52	52.22 ± 4.40	66.46 ± 5.50
Gender	Male	457 4(49.30%)	2559 (46.20%)	2015 (53.80%)
	Female	4707 (50.70%)	2976 (53.80%)	1731 (46.20%)
Smoking	Yes	3734 (40.20%)	2074 (37.50%)	1660 (44.30%)
	No	5547 (59.80%)	3461 (62.50%)	2086 (55.70%)
Drinking	Drink more than once a month	2461 (26.50%)	1461 (26.40%)	1000 (26.70%)
	Drink but less than once a month	742 (8.00%)	487 (8.80%)	255 (6.80%)
	None of these	6078 (65.50%)	3587 (64.80%)	2491 (66.50%)
Education	No formal education (illiterate)	2098 (22.60%)	1047 (18.90%)	1051 (28.10%)
	Elementary school or below	1718 (18.50%)	895 (16.20)	823 (21.90%)
	Middle school school and above	5465 (58.90%)	3593 (64.8%)	1872 (50.1%)
Marital status	Married with spouse present	7998 (86.20%)	4959 (89.60%)	3039 (81.10%)
	Married but not living with spouse temporarily for reasons such as work	385 (4.10%)	291 (5.30%)	94 (2.50%)
	Separated	29 (0.30%)	12 (0.20%)	17 (0.50%)
	Divorced	62 (0.70%)	38 (0.70%)	24 (0.60%)
	Widowed	741 (8.00%)	198 (3.60%)	543 (14.50%)
	Never married	66 (0.70%)	37 (0.70%)	29 (0.80%)
FIT1		0.09 ± 0.07	0.08 ± 0.07	0.11 ± 0.08
FIT2		0.07 ± 0.05	0.07 ± 0.05	0.08 ± 0.06
FIT3		0.07 ± 0.06	0.06 ± 0.05	0.08 ± 0.07
DST1		8.11 ± 6.22	7.77 ± 6.13	8.62 ± 6.31
DST2		7.75 ± 5.74	7.67 ± 5.75	7.87 ± 5.72
DST3		8.06 ± 6.42	7.84 ± 6.33	8.39 ± 6.53

3.2 Correlation analysis between frailty and depressive symptoms

The results of the correlation analysis were shown in Fig. 2. There was a significant positive correlation between frailty and depressive symptoms at each time point in the total sample (all P < 0.001), and a significant positive correlation between frailty and depressive symptoms at each time in both middle-aged and elderly groups at each time (all P < 0.001) (Supplementary Fig. 1&2).

3.3 Developmental trajectory of frailty and depressive symptoms

An unconditional linear growth model for the debilitation of the total sample population was first constructed to test the developmental trajectory of the frailty. Its model fitting results are not very satisfactory. The starting degree of fading was unrelated to its rate of change during the three measurements, which revealed a significant decreasing trend in fading (β = -0.007, P < 0.001). Unconditional linear growth models for frailty in the middle-aged and elderly populations were constructed separately by subgroups, and the results were consistent with those of the total sample (Table 2).

Table 2

Unconditional linear growth model of debilitating and depressive symptoms
Unconditional liner growth model	Model Fit Indices						Intercept			Slope			Slope with intercept
Unconditional liner growth model	χ²	df	CFI	TLI	RMSEA	SRMR	β	se	P	β	se	P	β	se	P
Total
Frailty	805.986	1	0.769	0.307	0.295	0.082	0.08	0.001	0	-0.007	0	0	0	0	0
Depressive symptoms	43.767	1	0.991	0.974	0.068	0.016	7.96	0.06	0	-0.021	0.033	0.529	1.338	0.414	0.001
45–59 years
Frailty	369.856	1	0.808	0.425	0.258	0.071	0.07	0.001	0	-0.007	0	0	0	0	0
Depressive symptoms	4.309	1	0.999	0.996	0.024	0.006	7.712	0.077	0	0.036	0.042	0.397	1.344	0.53	0.011
≥ 60 years
Frailty	423.824	1	0.708	0.125	0.336	0.097	0.096	0.001	0	-0.007	0.001	0	0.001	0	0
Depressive symptoms	62.855	1	0.972	0.916	0.128	0.029	8.324	0.0928	0	-0.113	0.052	0.03	1.221	0.662	0.065

Following that, a well-fitting unconditional linear growth model for depressive symptoms in the entire sample group was developed. The initial level of the intercept was 7.96. Levels of depressive symptoms were stable across the three measurements (β = − 0.021, P = 0.529). There was a significant positive correlation between the intercept and slope (β = 1.338, P = 0.001), which suggested that the higher the initial level of depressive symptoms in an individual, the faster their depression level grew. Unconditional linear growth models of depressive symptoms were constructed separately for the middle-aged and elderly populations, and the results were consistent with those for the total sample. However, the model results for depressive symptoms in the older population showed no association between intercept and slope (β = 1.221, P = 0.065), suggesting that in the older population, individuals' initial depressive symptom levels were independent of their rate of change (Table 2)

3.4 A Parallel Latent growth model for frailty and depressive symptoms

In order to investigate the relationship between the initial level of depressive symptoms, the development trajectory of depressive symptoms, the initial level of frailty and the development trajectory of weakness, first we constructed a parallel latent variable growth model containing the total sample of depressive symptoms and frailty. Model fitting indexes: χ² = 1090.078, df = 6, P < 0.05, CFI = 0.912, TLI = 0.781, AIC = 90155.755, BIC = 90305.605, RMSEA = 0.140. The results suggested that the model fitting was acceptable. The correlation coefficient between the intercept factor and slope factor for depressive symptoms was − 0.568 (P = 0.149), suggesting that the initial level of depressive symptoms was not associated with subsequent rates of development. The regression coefficients from the intercept factor to the weak intercept factor and slope factor of depressive symptoms were 0.009 (p < 0.001) and − 0.001 (P < 0.001), respectively, indicating that the initial level of depressive symptoms had a significant positive predictive effect on the initial level of frailty, but a significant negative predictive effect on the rate of change of frailty. The regression coefficients of the slope factor of depressive symptoms on the intercept factor and slope factor of frailty were − 0.023 (P < 0.001) and 0.020 (P < 0.001), and the results showed that the rate of change of depressive symptoms had a significant negative predictive effect on the initial level of frailty, but a significant positive predictive effect on the rate of change of frailty. The regression coefficient of the slope factor for the weakened intercept factor was − 0.001 (P < 0.001), indicating that there was a negative association between the initial level of weakness and the rate of development (Fig. 3).

Then we constructed parallel latent growth models for the middle-aged and elderly populations separately. A model for the middle-aged population is presented in the annex. Model fit metrics: χ² = 470.531, df = 6, P < 0.05, CFI = 0.933, TLI = 0.833, AIC = 49493.065, BIC = 49632.060, RMSEA = 0.118, SRMR = 0.044, and the results suggested that the model fitting results were acceptable. The model showed mostly consistent results with those for the total sample. Then, an unconditional linear growth model was developed and found to be well-fitting for depressive symptoms throughout the entire sample population (Supplementary Fig. 3). Models for the elderly population are presented in the Appendix, Model fitting indexes: χ² = 621.229, df = 6, P < 0.05, CFI = 0.881, TLI = 0.703, AIC = 39271.035, BIC = 39401.832, RMSEA = 0.165, SRMR = 0.061. The results suggested that the model fitting results were acceptable. The model showed mostly consistent results with those for the total sample. Furthermore, the intercept component for depressive symptoms on the slope factor for frailty in this model had a regression coefficient of 0 (P = 0.275), which indicated that the initial level of depressive symptoms was not linked to the rate of frailty development in the senior population. In this model, the baseline degree of frailty and the rate of development exhibited a positive correlation, which was in line with the findings for the middle-aged population (Supplementary Fig. 4).

3.5 Cross-lagged panel analysis

Figure 4 showed a cross-lagged model built at three-time points: frailty and depressive symptoms, which fits well (CFI = 0.938, TLI = 0.766, RMSEA = 0.182, SRMR = 0.052). Cross-lagged path analysis showed significant correlations between frailty and depressive symptoms at T1, T2, and T3. Depressive symptoms (β = 7.32, P < 0.001) and frailty (β = 0.39, P < 0.001) at moment T2 were found to significantly predict by frailty at moment T1. Depressive symptoms at T1 significantly predicted frailty (β = 0.001, P < 0.001) and depressive symptoms (β = 0.43, P < 0.001) at T2. Frailty at T2 significantly predicted depressive symptoms (β = 17.44, P < 0.001) and frailty (β = 0.69, P < 0.001) at T3. Depressive symptoms at T2 significantly predicted frailty (β = 0.001, P < 0.001) and depressive symptoms (β = 0.55, P < 0.001) at T3.

Cross-lagged models were constructed to investigate the interpredictive relationship between frailty and depressive symptoms by grouping the total sample by age. The model fit was good for the middle-aged group (45–59 years) (CFI = 0.936, TLI = 0.762, RMSEA = 0.181, SRMR = 0.053). Model fit was acceptable for the older group (≥ 60 years) (CFI = 0.944, TLI = 0.789, RMSEA = 0.171, SRMR = 0.051). The outcomes of the cross-lagged model in the categories of middle-aged and old people were in line with the outcomes of the cross lag model in the entire sample. Therefore, the cross-lagged model for the total sample was taken as the model investigated in this study (Supplementary Fig. 5&6).

In this study, we investigated the changes of frailty and depressive symptoms in the middle-aged and elderly Chinese population, as well as the interaction between frailty and depressive symptoms. The main findings suggested that, first, the higher the initial level of depressive symptoms, the slower the rate of decline changed. Second, the faster the rate of change in depressive symptoms, the faster the rate of changed in frailty. Third, there was a bidirectional causal relationship between debilitation and depressive symptoms. Finally, the bidirectional causality between frailty and depressive symptoms was independent of age.

In both the total sample population and the middle-aged population, it was discovered that the pace of change in frailty was slower the more severe the starting level of depressive symptoms were. This may be because the higher the initial level of depressive symptoms, the higher the initial level of frailty, and therefore its rate of change becomes slow. We observed that in the elderly population, the initial severity of depressive symptoms was not associated with the rate of frailty development. This finding may be explained by the elderly population's increased risk of accidents, such as falls, fractures, etc., as well as their consistent decline in activity, both of which are frailty components. As a result, the rate of frailty development in the elderly population may be primarily related to these factors. Thus, the effect of initial levels of depressive symptoms was relatively reduced to none. In this investigation, we discovered that the rate of change in frailty in the entire sample population, the middle-aged population, and the senior population was inversely correlated with the rate of change in depressive symptoms. The reason for this may be that depressive symptoms are positively correlated with frailty, and the level of frailty increases as depressive symptoms increase.

The findings of this study were consistent with a number of earlier findings, which indicated that frailty in the middle-aged and elderly Chinese population predicted the development of depressive symptoms^{[17, 31]}. Meanwhile, frailty was found to be a predictor of depressive symptoms in a cross-sectional research of 576 older persons in an urban community in Shenzhen, China, who were 65 years of age or older^[32]. A possible explanation for this relationship was that when older adults suffer from the debilitating syndrome, they had low physical activity, slow movement, fatigue and weakness, and may refuse social activities, and after many times, older adults may develop feelings of uselessness and emptiness, no longer be interested in what was originally interesting, and then depressive symptoms. Another possibility is that interleukin-6 (IL-6) levels are higher in fragile elderly persons^[33], which, as a biomarker of depressive symptoms in older patients, is thought to be more susceptible to geriatric depression. Thus, frailty was a predictor of depressive symptoms.

In the Chinese middle-aged and elderly population, a striking finding of this study was that depressive symptoms were a predictor of frailty. This finding was inconsistent with a previous longitudinal study^[17], and differences in results may be due to different methods of assessing frailty. However, another cross-sectional study of 5844 participants from seven cities in China showed that depression was a risk factor for frailty^[12]. As well, using data from the Rugao Longevity and Aging Study (RuLAS), an analysis of 1168 Chinese older adults aged 70 years and older, Zhang et al. found that depressive symptoms were a risk factor for frailty^[34]. In addition, in studies of older populations from Brazil and Latin America, the results also showed that depression increased the risk of frailty^{[11, 35]}. Therefore, depressive symptoms may also increase the risk of debilitation in the older Chinese population. One possible explanation was that depression causes depressed mood, weight loss and loss of appetite in older adults, which in turn triggers symptoms such as reduced mobility, weakness and falls, increasing their risk of developing frailty.

Our study identifies a meaningful relationship with a bidirectional causal association between frailty and depressive symptoms in the Chinese middle-aged and elderly population. This is in accordance with the findings of a previous study of older adults in rural China^[16]. In addition, a cross-lagged model analysis using data from the CHARLS database also found a bidirectional relationship between loneliness and frailty^[36]. While loneliness was measured by a question on the CESD-10 scale in that study, there may be a bidirectional association between frailty and depressive symptoms among Chinese older adults. Additionally, our research revealed that in middle-aged populations, there is a bidirectional causal relationship between frailty and depressive symptoms that is independent of age. This meant that early recognition and intervention that also focuses on depressive symptoms and frailty in middle-aged populations should be considered to delay further deterioration upon entry into old age.

China is currently experiencing deep aging, with a sizable section of the population being elderly, and the impact of the elderly's frailty and depression on society and public health is tremendous. This study established a bidirectional causal relationship between frailty and depressive symptoms, which could delay the development of depression and frailty by identifying and intervening in modifiable indicators of depressive symptoms in frailty middle-aged and elderly or modifiable indicators of frailty in middle-aged and elderly patients with depressive symptoms, and can improve the physical health and mental health of elderly people, thereby promoting healthy aging.

The following are the study's advantages: First, this study used a representative sample of Chinese countries, thus the study results can be generalized to the whole country. Secondly, the parallel latent growth model was used to reveal the development trajectory and the interaction between them from a dynamic perspective. Thirdly, this study used cross-lagged path analysis, which can reliably investigate the bidirectional predictive relationship between frailty and depressive symptoms. The present study, however, also has several limitations: Firstly, the depressive symptom assessment in this study was self-reported; Second, in this cohort study, we eliminated missing values and did not use some effective methods for imputation, which may cause some deviations in the results; Third, the path analysis in this study was assuming that there was a linear correlation between frailty and depressive symptoms, however, the relationship between variables was subtle, such as a "U" -type relationship. Therefore, further prospective studies are needed to validate our results.

In summary, using longitudinal data from the CHARLS database, our findings suggest taking into account the existence of longitudinal and reciprocal relationships between frailty and depressive symptoms through latent growth models and cross-lagged analyses. Health care providers can devote more interventions to any symptom of frailty or depressive symptoms, which can delay the course of another symptom and achieve a doubling of effect.

Ethics approval and consent to participate.

All methods in this study on humans described in the manuscript were performed in accordance with national law and the Helsinki Declaration of 1975 and its later amendments. The study was approved by the Biomedical Ethics Review Committee of Peking University (IRB00001052–11015). All respondents were informed about the purpose of the study and consented to participate.

Consent for publication

Not applicable.

Availability of Data and Materials

Dataset from the China Health and Retirement Longitudinal Study (CHARLS) http://charls.pku.edu.cn/.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the Inflammation and Immune Mediated Diseases Laboratory of Anhui Province (Hefei, 230032, IMMDL202109), scientific research cultivation project for School of Nursing, Anhui Medical University (hlpy20210011), Graduate green seedling cultivation project for School of Nursing, Anhui Medical University (hlqm12023018) and the Research Fund of Anhui nursing Association (AHHL a202112).

Authors' contributions

Conceptualization: Wu GC and Pan HF. Formal analysis and Original draft preparation: Sang N. Methodology: Sang N, Liu RC and Zhang MY. Data curation: Lu ZX and Zhang MH. Review and editing: Li BH and Wei M. All authors have read and approved the final version of the manuscript.

Acknowledgements

We thank the China Health and Retirement Longitudinal Study (CHARLS) team for providing data.

WHO. 2015. China country assessment report on ageing and health. Switzerland: WHO Press.
Statistics NBo. 2022. Statistical Bulletin of National Economic and Social Development of the People 's Republic of China 2022.
Dent E, Martin FC, Bergman H, Woo J, Romero-Ortuno R, Walston JD (2019). Management of frailty: opportunities, challenges, and future directions. Lancet, 394:1376-1386.
Hoogendijk EO, Afilalo J, Ensrud KE, Kowal P, Onder G, Fried LP (2019). Frailty: implications for clinical practice and public health. Lancet, 394:1365-1375.
Wilkinson P, Ruane C, Tempest K (2018). Depression in older adults. Bmj, 363:k4922.
Alexopoulos GS (2005). Depression in the elderly. Lancet, 365:1961-1970.
Lohman M, Dumenci L, Mezuk B (2014). Sex differences in the construct overlap of frailty and depression: evidence from the Health and Retirement Study. J Am Geriatr Soc, 62:500-505.
Soysal P, Veronese N, Thompson T, Kahl KG, Fernandes BS, Prina AM, et al. (2017). Relationship between depression and frailty in older adults: A systematic review and meta-analysis. Ageing Res Rev, 36:78-87.
Vaughan L, Corbin AL, Goveas JS (2015). Depression and frailty in later life: a systematic review. Clin Interv Aging, 10:1947-1958.
Makizako H, Shimada H, Doi T, Yoshida D, Anan Y, Tsutsumimoto K, et al. (2015). Physical frailty predicts incident depressive symptoms in elderly people: prospective findings from the Obu Study of Health Promotion for the Elderly. J Am Med Dir Assoc, 16:194-199.
Borges MK, Romanini CV, Lima NA, Petrella M, da Costa DL, An VN, et al. (2021). Longitudinal Association between Late-Life Depression (LLD) and Frailty: Findings from a Prospective Cohort Study (MiMiCS-FRAIL). J Nutr Health Aging, 25:895-902.
Ma L, Tang Z, Zhang L, Sun F, Li Y, Chan P (2018). Prevalence of Frailty and Associated Factors in the Community-Dwelling Population of China. J Am Geriatr Soc, 66:559-564.
Qiu QW, Qian S, Li JY, Jia RX, Wang YQ, Xu Y (2020). Risk factors for depressive symptoms among older Chinese adults: A meta-analysis. J Affect Disord, 277:341-346.
Zhang J, Xu L (2022). Frailty and Associated Factors among Chinese Centenarians. J Nutr Health Aging, 26:806-813.
Ferrucci L, Fabbri E (2018). Inflammageing: chronic inflammation in ageing, cardiovascular disease, and frailty. Nat Rev Cardiol, 15:505-522.
Wang X, Shen K (2021). The Reciprocal Relationship between Frailty and Depressive Symptoms among Older Adults in Rural China: A Cross-Lag Analysis. Healthcare (Basel), 9.
Cao L, Zhou Y, Liu H, Shi M, Wei Y, Xia Y (2022). Bidirectional Longitudinal Study of Frailty and Depressive Symptoms Among Older Chinese Adults. Front Aging Neurosci, 14:791971.
Chen X, Mao G, Leng SX (2014). Frailty syndrome: an overview. Clin Interv Aging, 9:433-441.
Gordon SJ, Baker N, Kidd M, Maeder A, Grimmer KA (2020). Pre-frailty factors in community-dwelling 40-75 year olds: opportunities for successful ageing. BMC Geriatr, 20:96.
Jiang R, Noble S, Sui J, Yoo K, Rosenblatt M, Horien C, et al. (2023). Associations of physical frailty with health outcomes and brain structure in 483 033 middle-aged and older adults: a population-based study from the UK Biobank. Lancet Digit Health, 5:e350-e359.
Zhao Y, Hu Y, Smith JP, Strauss J, Yang G (2014). Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol, 43:61-68.
Rockwood K, Song X, MacKnight C, Bergman H, Hogan DB, McDowell I, et al. (2005). A global clinical measure of fitness and frailty in elderly people. Cmaj, 173:489-495.
Searle SD, Mitnitski A, Gahbauer EA, Gill TM, Rockwood K (2008). A standard procedure for creating a frailty index. BMC Geriatr, 8:24.
Liu X, Dai G, He Q, Ma H, Hu H (2022). Frailty Index and Cardiovascular Disease among Middle-Aged and Older Chinese Adults: A Nationally Representative Cross-Sectional and Follow-Up Study. J Cardiovasc Dev Dis, 9.
He D, Qiu Y, Yan M, Zhou T, Cheng Z, Li J, et al. (2023). Associations of metabolic heterogeneity of obesity with frailty progression: Results from two prospective cohorts. J Cachexia Sarcopenia Muscle, 14:632-641.
Chen H, Mui AC (2014). Factorial validity of the Center for Epidemiologic Studies Depression Scale short form in older population in China. Int Psychogeriatr, 26:49-57.
Meredith W (1990). Latent curve analysis. Psychometrika:107-122.
Little TD. 2013. Longitudinal structural equation modeling.
Hamaker EL, Kuiper RM, Grasman RP (2015). A critique of the cross-lagged panel model. Psychol Methods, 20:102-116.
Hu Lt, Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6:1-55.
Liu H, Li D, Zhao X, Fang B, Zhang Q, Li T (2021). Longitudinal Impact of Frailty States and Sleep Duration on Subsequent Depressive Symptoms of Older Adults. J Am Geriatr Soc, 69:1003-1011.
Wang Q, Song D, Lin Q, Tao A, Zhang Y, Li X, et al. (2023). The critical role of physical frailty and function on depressive symptoms among community-dwelling older adults in China: A cross-sectional study. Front Psychiatry, 14:1134427.
Xu Y, Wang M, Chen D, Jiang X, Xiong Z (2022). Inflammatory biomarkers in older adults with frailty: a systematic review and meta-analysis of cross-sectional studies. Aging Clin Exp Res, 34:971-987.
Zhang N, Shi GP, Wang Y, Chu XF, Wang ZD, Shi JM, et al. (2020). Depressive symptoms are associated with incident frailty in a Chinese population: the Rugao Longevity and Aging Study. Aging Clin Exp Res, 32:2297-2302.
Prina AM, Stubbs B, Veronese N, Guerra M, Kralj C, Llibre Rodriguez JJ, et al. (2019). Depression and Incidence of Frailty in Older People From Six Latin American Countries. Am J Geriatr Psychiatry, 27:1072-1079.
Sha S, Pan Y, Xu Y, Chen L (2022). Associations between loneliness and frailty among older adults: Evidence from the China Health and Retirement Longitudinal Study. BMC Geriatr, 22:537.

No competing interests reported.

Supplementarymaterials.docx

Download PDF

Journal Publication

published 25 Jan, 2024

Read the published version in BMC Public Health →

Editorial decision: Revision requested
24 Nov, 2023
Reviews received at journal
22 Oct, 2023
Reviewers agreed at journal
22 Oct, 2023
Reviewers invited by journal
11 Oct, 2023
Editor assigned by journal
11 Oct, 2023
Editor invited by journal
31 Aug, 2023
Submission checks completed at journal
31 Aug, 2023
First submitted to journal
30 Jul, 2023

You are reading this latest preprint version

Changes in frailty and depressive symptoms among middle-aged and elderly Chinese: a nationwide cohort study

Status:

Journal Publication

Version 1

Abstract

Background and Aims:

Methods

Results

Conclusions

Figures

1 Introduction

2 Methods

2.1 Data Source and Participants

2.2 Measurements

2.2.1 Frailty

2.2.2 Depressive symptoms

2.2.3 Demographic variables

2.4 Statistical Analysis

3 Results

3.1 Descriptive Analysis

3.2 Correlation analysis between frailty and depressive symptoms

3.3 Developmental trajectory of frailty and depressive symptoms

3.4 A Parallel Latent growth model for frailty and depressive symptoms

3.5 Cross-lagged panel analysis

4 Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1