Bidirectional Association between Sarcopenia and Depressive Symptoms among Chinese Middle- and Older-Aged Adults: Longitudinal Observational Study

Background: The study aimed to examine the bidirectional relationship between sarcopenia and depressive symptoms in a national, community-based cohort study, despite the unclear temporal sequence demonstrated previously. Methods: Data were derived from four waves (2011 baseline and 2013, 2015, and 2018 follow-ups) of the China Health and Retirement Longitudinal Study (CHARLS). A total of 17,708 participants aged 45 years or older who had baseline data on both sarcopenia status and depressive symptoms in 2011 were included in the study. For the two cohort analyses, a total of 8092 adults without depressive symptoms and 11,292 participants without sarcopenia in 2011 were included. Sarcopenia status was defined according to the Asian Working Group for Sarcopenia 2019 (AWGS 2019) criteria. Depressive symptoms were defined as a score of 20 or higher on the 10-item Center for Epidemiologic Studies Depressive Scale (CES-D-10). Cox proportional hazard regression models were conducted to examine the risk of depressive symptoms and sarcopenia risk, while cross-lagged panel models were used to examine the temporal sequence between depressive symptoms and sarcopenia over time. Results: During a total of 48,305.1 person-years follow-up, 1262 cases of incident depressive symptoms were identified. Sarcopenia exhibited a dose–response relationship with a higher risk of depressive symptoms (HR = 1.7, 95%CI: 1.2–2.3 for sarcopenia, and HR = 1.5, 95%CI: 1.2–1.8 for possible sarcopenia, p trend < 0.001). In the second cohort analysis, 240 incident sarcopenia cases were identified over 39,621.1 person-years. Depressive symptoms (HR = 1.5, 95%CI: 1.2–2.0) are significantly associated with a higher risk of developing sarcopenia after multivariable adjustment (p < 0.001, Cross-lagged panel analyses demonstrated that depressive symptoms were associated with subsequent sarcopenia (β = 0.003, p < 0.001). Simultaneously, baseline sarcopenia was also associated with subsequent depressive symptoms (β = 0.428, p < 0.001). Conclusion: This study identified a bidirectional relationship between depressive symptoms and sarcopenia. It seems more probable that baseline sarcopenia is associated with subsequent depressive symptoms in a stronger pattern than the reverse pathway. The interlinkage indicated that maintaining normal muscle mass and strength may serve as a crucial intervention strategy for alleviating mood disorders.


Introduction
Sarcopenia and depressive symptoms are two major contributors to disease burden worldwide, limiting functional abilities and diminishing quality of life in the elderly [1][2][3][4].Both conditions were common, with major depressive disorder having a life-time prevalence of an estimated 15.0-18.0%and 4.6-9.3% in the elderly population [5][6][7][8].Conversely, the prevalence of sarcopenia ranged from 8% to 36% in middle-aged and 11.5% to 40% in older adults [9][10][11][12].Both depressive symptoms and sarcopenia are associated with a substantial risk of negative outcomes, resulting in personal suffering, family disruption, and imposing a substantial economic burden on society [13][14][15].Previous studies have reported that these two conditions have the potential to engender significant health burdens and considerable adverse consequences in conjunction with the global aging population trend [16,17].
According to previous studies, adults with sarcopenia and depressive symptoms had lower cognitive performance [24,25] and physical activity levels [26,27] and higher risk of falls [28,29] than those without these conditions.Other studies revealed that older adults with both sarcopenia and depressive symptoms had a higher risk of falls and hospitalization than those without these conditions [18].Additionally, it has been reported that both sarcopenia and depressive symptoms serve as risk factors for cardiovascular disease and are linked to long-term mortality associated with it [30][31][32].In summary, the co-occurrence of the two conditions results in impaired functioning and incapacity, which diminishes overall well-being and raises the risk of illness and death.
However, while the majority of previous studies primarily focused on establishing a correlation between sarcopenia and depressive symptoms using cross-sectional data [19,33], more recent longitudinal studies have begun to explore the prospective effects of depression on sarcopenia.Nevertheless, the bidirectional and temporal sequence between sarcopenia and depressive symptoms remains unclear.Understanding the potential temporal sequence between sarcopenia and depressive symptoms is crucial for early detection and intervention, ultimately facilitating effective disease management and improving overall quality of life.Thus, to address these knowledge gaps, our study aimed to investigate the bidirectional association between sarcopenia and depressive symptoms using data from a large Chinese national community-based prospective cohort study with an extensive long-term follow-up.

Study Population
All of data in this study were collected from the 2011, 2013, 2015, and 2018 waves of the China Health and Retirement Longitudinal Study (CHARLS), including the assessment of depression and sarcopenia.CHARLS constitutes a continuous survey program characterized by periodic examinations recurring at intervals of 2 to 3 years.Within this framework, participants undergo face-to-face interviews within the confines of their own residences, facilitated by the deployment of computer-assisted personal interviewing (CAPI) technology.Detailed information regarding the study design of the CHARLS have been previously reported [34].The Institute of Social Science Survey at Peking University maintains all CHARLS-collected data, which has been made available to the public via the CHARLS website (http://charls.pku.edu.cn/en,accessed on 27 May 2023).In this present study, we analyzed data from the CHARLS 2011, 2013, 2015, and 2018.A total of 17,708 respondents were recruited in the 2011 survey.The inclusion criteria for the cohort 1 (cohort for the onset of depressive symptoms) were as follows: (1) individuals who were 45 years of age or older in 2011; (2) individuals who had baseline data available for assessing sarcopenia status; (3) individuals who neither received a diagnosis of mood disorders from medical professionals nor took any medication for addressing emotional, nervous, or mental issues; and (4) those who obtained CES-D-10 scores ≤ 10 in 2011.The exclusion criteria were as follows: (1) individuals who did not engage in the assessment of depressive symptoms in 2013, 2015, or 2018; and (2) individuals with missing data on all demographic characteristics and confounders.Ultimately, 8092 participants were included in the first cohort analysis.For cohort 2, focused on the onset of sarcopenia, inclusion criteria were (1) individuals aged 45 years or older in 2011; and (2) individuals who had baseline data on the assessment of depressive symptoms in 2011, either through participation in the CES-D-10 evaluation or by reporting a diagnosis of mood disorders by doctors or taking medication for mood disorders.The exclusion criteria comprised the following: (1) absence of data pertaining to all components associated with sarcopenia in the years 2013, 2015, or 2018; (2) participants who already had sarcopenia at baseline; (3) lack of information concerning demographic variables and covariates.Finally, 11,292 participants were included in the second cohort analysis (Table A1).Within cohort 1, approximately half of the participants were female (3717/8092, 45.93%), with a median age of 57.0 years.Cohort 2 comprised 47.33% females (5345/11,292), with a median age of 55.0 years.

Assessment of Sarcopenia
The assessment of sarcopenia was conducted in accordance with the 2019 Asian Working Group for Sarcopenia (AWGS) criteria, which included evaluations of physical performance, appendicular skeletal muscle mass (ASM), and muscle strength.The hands of both the dominant and non-dominant individuals were assessed for handgrip strength, which was quantified in kilograms, using a Yuejian TM WL-1000 dynamometer (Nantong, China).Participants were instructed to exert maximal force while squeezing the dynamometer, with two measurements taken for each hand while holding the dynamometer at a right angle (90 • ).In analyses, the mean value of the maximal strength data was utilized.When measurement of one hand was impossible because of any reason, the greatest score obtained from the other hand was documented.Low handgrip strength was defined as less than 28 kg for males and less than 18 kg for females, in accordance with the AWGS 2019 criteria.A validated anthropometric equation was employed to estimate the ASM of the Chinese population, as reported by Chinese residents [30,35].
The weight and height of the subjects were determined utilizing an Omron TM HN-286 scale (Kyoto, Japan) and a Seca TM 213 height meter (Hamburg, Germany), respectively.A former study has illustrated that the ASM computed through this formula strongly correlates with measurements derived from dual-energy X-ray absorptiometry (DXA) [36].The threshold for defining low muscle mass was determined by identifying the sex-specific lowest 20th percentile of height-adjusted muscle mass (ASM/height 2 ) within the study cohort, with values of <5.29 kg/m 2 for women and <6.99 kg/m 2 for men, consistent with previous studies [37].
Physical performance evaluation encompassed measures of gait speed, the five-time chair stand test, and the Short Physical Performance Battery (SPPB) [38].In order to evaluate gait speed, participants were directed to walk a distance of 2.5 m at their typical speed, and their durations of completion were documented over the course of two trials.The five-time chair stand test documented the amount of time necessary to ascend consecutively from a chair with a cross-chest fold, measuring 47 cm in height.The SPPB assessment consisted of three balance tests, each lasting 10 s: (1) side-by-side stance; (2) semi-tandem stance (with the heel of one foot placed next to the big toe of the other foot); and (3) tandem stance (with the heel of one foot contacting the toes of the other foot).Each person's examination was given a maximum of four points, resulting in a cumulative score of twelve.Per the AWGS 2019 guidelines, low physical performance was characterized by a gait speed below 1.0 m/s, a five-time chair stand test taking 12 s or longer to complete, or an SPPB score falling below 9.
Possible sarcopenia can be identified by reduced muscular strength or poor physical performance.Sarcopenia is characterized by an accompanying reduction in muscle mass, strength, or physical performance.Consequently, participants were categorized into three groups: non-sarcopenia, possible sarcopenia, and sarcopenia.

Depressive Symptoms Ascertainment
This study assessed depressive symptoms using the 10-item Center for Epidemiologic Studies Depressive Scale (CES-D-10), which is a refined version derived from the original 20-item CES-D.Participants were queried about the frequency of experiencing ten listed symptoms during the preceding week.Responses were categorized as follows: rarely or never (<1 day), sometimes (1-2 days), occasionally (3-4 days), and most or all of the time (5-7 days).Each item was assigned a score from 0 (rarely) to 3 (most or all of the time), with items 5 and 8 reverse-scored before score summation.The total CES-D-10 score ranged from 0 to 30.Consistent with prior research, participants scoring 20 or higher were classified into the depressive symptoms group, those with a score of ≥10 were categorized similarly, while those scoring <10 were designated as the no depressive symptoms group.[39,40].The CES-D-10 has undergone comprehensive validation for application in general populations and has exhibited sufficient levels of reliability and validity when applied to senior adults residing in the community in China [41].

Covariates
Covariates were chosen in consideration of prior epidemiological evidence and the availability of data at baseline [10,17,24,30].Potential confounding variables encompassed age (continuous), sex (male and female), BMI (<25 kg/m 2 and ≥25 kg/m 2 ), residential location (urban and rural), educational attainment (from illiterate to primary school, middle school, and high school to college or higher enrolled), average household income (CNY <1000, 1000, 5000, 10,000, and 20,000), marital status (including married, separated, and unmarried/divorced/widowed), smoking status (including never, current, and previous), alcohol intake (including never, current, and previous), daily sleep time, presence of comorbid physical diseases (including cancer, chronic lung diseases, heart disease, stroke, arthritis, dyslipidaemia, hepatic disease, kidney disease, digestive system disease, asthma, and memory-related diseases such as dementia, brain atrophy, and Parkinson's disease, as well as hypertension and diabetes), and cognitive function.The presence of comorbidities was assessed through the CHARLS questionnaire, which inquired about participants' diagnosed conditions and treatments by asking, "Have you been diagnosed with this condition by a doctor?" and "Are you currently undergoing any treatment for this condition?"Participants who affirmed either having the condition or undergoing treatment were categorized as having the condition in our study.Cognitive function was assessed by Mini-Mental State Examination (MMSE) results [42].

Statistical Analysis
In order to summarize continuous variables, means (standard deviations (SDs)) or medians (interquartile range (IQR)) were utilized.Subsequently, to compare differences between groups, we conducted independent t tests, one-way analysis of variance (ANOVA) analyses, Wilcoxon tests, and Kruskal-Wallis tests.χ 2 tests were performed to compare categorical or ordinal variables represented as numbers (proportions).The cumulative incidence rate was determined using the Kaplan-Meier method.Subsequently, a Cox proportional hazards model was employed to investigate the relationship between sarcopenia and depressive symptoms.The follow-up duration started from the baseline assessment in 2011 and continued to the occurrence of the first reported outcome (sarcopenia or depressive symptoms).For participants who did not develop sarcopenia or depressive symptoms during the study period, the follow-up duration was determined from the baseline to the censoring date at the end of the study (2018), the date of death, or to the loss of follow-up.
In order to explore the reciprocal relationship between sarcopenia and depressive symptoms, three models were performed: (1) model 1, adjusted for age and sex; (2) model 2, additionally adjusted for residential region, education level, average household income, BMI, marital status, status of smoking, and alcohol intake; (3) model 3, including comorbidities, sleep duration, and MMSE score.A trend test was conducted by allocating a score value (0, 1, 2) to each sarcopenia status (non-sarcopenia, possible sarcopenia, and sarcopenia) and modeling this value as a continuous variable.The Wald test was then used to determine the statistical significance of this value.
Furthermore, a subgroup analysis was conducted to examine potential variations in the association between sarcopenia and depressive symptoms based on age, gender, sleep duration (≤6 h, >6 h), and physical disease comorbidity.In order to examine effect modification, interaction terms between stratified variables were incorporated.In the sensitivity analysis, participants who disclosed incident outcomes within a two-year period following recruitment were removed.Second, a competing risk model was implemented, which incorporated mortality and loss-of-follow-up as competing events, given that the participants in that model might subsequently develop outcomes.
Participants from the prospective cohort analysis who had comprehensive data on both depressive symptoms and sarcopenia at baseline (time 1) and who also completed at least one follow-up assessment were included in the cross-lagged path analysis.At the follow-up (time 2), depressive symptoms and sarcopenia were assessed using data collected from the participant with the longest duration of complete follow-up.Before conducting the cross-lagged path analysis, regression residual analyses were employed to modify the baseline and subsequent values of depressive symptoms and sarcopenia status for a set of covariates.Subsequently, the values were standardized using Z-transformation (mean 0, SD 1).Coefficients that were standardized were documented for each model.Age, gender, location of residence, marital status, level of education, smoking, and alcohol consumption have been adjusted for in all models.To demonstrate that the models were well-fitting, a comparative fit index (CFI) greater than 0.90 and a standardized root-mean-square residual (SRMR) less than 0.05 were used [43].Analyses were conducted using SAS software version 9.4 and R version 4.0.2, with a p value < 0.05 indicating statistical significance.

Results
A total of 25,586 participants were interviewed in CHARLS 2011.Subsets of 13,210 and 7791 participants were excluded from the analysis due to the absence of data pertaining to the assessment of sarcopenia or depressive symptoms at both the baseline and follow-up for the two cohorts.In the longitudinal analysis, an additional 4284 participants with a mood disorder diagnosed by a psychiatrist at baseline were excluded from the depressive symptoms cohort, and 6503 participants with sarcopenia at baseline were excluded from the sarcopenia cohort.Our final analytic sample consisted of 8092 and 11,292 participants for the two prospective cohorts, respectively (Figure 1).

Prospective Effect of Sarcopenia Status on the Onset of Depressive Symptoms
Among 8092 participants (median age 57.0, 51.0-64.0),54.1% were male.Overall, 2379 (29.4%) and 631 (7.9%) were possible sarcopenia and sarcopenia at baseline.There were 1205 (27.5%) cases of possible sarcopenia and 307 (7.0%) cases of sarcopenia in male participants, whereas there were 1174 (31.6%) cases of possible sarcopenia and 324 (8.7%) cases of sarcopenia in female participants.Participants with possible sarcopenia and sarcopenia were more likely to be female, live in a rural area, have a lower education level, have a lower level of household income, be divorced or separated, and have a lower BMI and lower cognitive score (Tables 1 and A2).The median follow-up duration of this study was 7.0 (interquartile range: 5.0-7.0)years.

Subgroup Analysis
The higher risk of depressive symptoms associated with possible sarcopenia and sarcopenia was observed in overall and in older participants; depressive symptoms were not associated with sarcopenia in the middle-aged group but were significantly associated with possible sarcopenia in this age group (Figure 3).Sarcopenia and possible sarcopenia were significantly associated with an increased risk of depressive symptoms in individuals who were male and older and had more than one type of physical illness.Female, middle-aged participants and participants without a physical illness were more likely to have possible sarcopenia than sarcopenia (Figure 3, Tables 2 and A3).Independently of participants' sleep duration, depressive symptoms were associated with sarcopenia and possible sarcopenia.On the other hand, the increased risk of sarcopenia associated with depressive symptoms was consistently observed across various subgroups, including age, sex, sleep duration, and comorbidities (Table 3, Figure 4).Furthermore, significant interaction across sex and depressive symptoms and was observed, with a higher risk of sarcopenia in females (HR = 1.2, 95%CI: 1.5-3.2, Figure 4).

Sensitivity Analysis
Results of sensitivity analysis by each single status of sarcopenia and each specific depressive symptoms were similar to primary findings (Figures A1 and A2).Sensitivity

Sensitivity Analysis
Results of sensitivity analysis by each single status of sarcopenia and each specific depressive symptoms were similar to primary findings (Figures A1 and A2).Sensitivity analysis excluding incident depressive symptoms cases within 2 years after baseline by performing a competing risk model and a further adjustment of major chronic physical diseases did not markedly affect the results, which indicated the robustness of our results (Table A4).

Bidirectional Association of Depressive Symptoms and Sarcopenia
Figure 5A illustrates the cross-lagged model exploring the bidirectional relationship between depressive symptoms and sarcopenia (model 1).After controlling for covariates, model 1 fitted the data adequately (CFI = 0.999; SRMR = 0.034).Depressive symptoms at time point 1 were positively related to themselves over time, as was sarcopenia.Furthermore, significant cross-lagged effects emerged, with preceding depressive symptoms significantly predicting sarcopenia at T2 (β = 0.003; p < 0.001) and vice versa, with sarcopenia predicting subsequent depressive symptoms at T2 (β = 0.428; p < 0.001).The standardized path coefficient from sarcopenia to depressive symptoms was significantly larger than that of the reverse pathway (χ 2 = 28.516,p for difference < 0.001).These findings provide empirical support for our hypothesis positing a bidirectional relationship between depressive symptoms and sarcopenia.
Figure 5B delineates the cross-lagged path analysis examining the interplay between sarcopenia and depressive symptoms within a middle-aged population (model 2).Notably, a significant cross-lagged pathway was observed.Specifically, the standardized path coefficient from sarcopenia at time 1 to depressive symptoms at time 2 (β = 0.411; p < 0.001) exhibited notable strength compared to the corresponding pathway from depressive symptoms at time 1 to sarcopenia at time 2 (β = 0.003; p < 0.001; χ 2 = 13.201,p for difference = 0.0002798).The fitting indicators (CFI = 0.999 and SRMR = 0.030) underscored the favorable model fit.Figure 5C illustrates the cross-lagged path analysis of sarcopenia and depressive symptoms in the older population (model 3).The cross-lagged pathway was evident, and the standardized path coefficient from sarcopenia at time 1 to depressive symptoms at time 2 (β = 0.486; p < 0.001) was markedly stronger than that from depressive symptoms at time 1 to sarcopenia at time 2 (β = 0.005; p < 0.001; χ 2 = 18.936, p for difference < 0.001).Fitting indicators (CFI = 0.995 and SRMR = 0.048) suggested model 3 to be a good model fit.
time point 1 were positively related to themselves over time, as was sarcopenia.Furthermore, significant cross-lagged effects emerged, with preceding depressive symptoms significantly predicting sarcopenia at T2 (β = 0.003; p < 0.001) and vice versa, with sarcopenia predicting subsequent depressive symptoms at T2 (β = 0.428; p < 0.001).The standardized path coefficient from sarcopenia to depressive symptoms was significantly larger than that of the reverse pathway (χ 2 = 28.516,p for difference < 0.001).These findings provide empirical support for our hypothesis positing a bidirectional relationship between depressive symptoms and sarcopenia.Figure 5B delineates the cross-lagged path analysis examining the interplay between sarcopenia and depressive symptoms within a middle-aged population (model 2).Notably, a significant cross-lagged pathway was observed.Specifically, the standardized path coefficient from sarcopenia at time 1 to depressive symptoms at time 2 (β = 0.411; p < 0.001) exhibited notable strength compared to the corresponding pathway from depressive symptoms at time 1 to sarcopenia at time 2 (β = 0.003; p < 0.001; χ 2 = 13.201,p for difference = 0.0002798).The fitting indicators (CFI = 0.999 and SRMR = 0.030) underscored the favorable model fit.Figure 5C illustrates the cross-lagged path analysis of sarcopenia and depressive symptoms in the older population (model 3).The cross-lagged pathway was ev-

Discussion
This study aimed to examine the bidirectional association between sarcopenia and depressive symptoms using data from a large Chinese national community-based prospective cohort study with an extensive long-term follow-up.Throughout the seven-year follow-up period, a longitudinal bidirectional association was observed between sarcopenia status and depressive symptoms among Chinese middle-aged and older adults.Our findings revealed a 50% and 70% elevated risk of depressive symptoms among participants with possible sarcopenia and sarcopenia, respectively, along with a 50% increased risk of sarcopenia in individuals experiencing depressive symptoms at the outset of the study.The increased risk of depressive symptoms was significantly identified in male, older participants with sarcopenia, whereas sarcopenia risk was higher in female and middle-aged individuals who experienced depressive symptoms.In addition, an examination of standardized cross-lagged coefficients indicated that the temporal connection between initial depressive symptoms and sarcopenia may be more pronounced than the inverse relationship.These results highlight the intricate interplay between sarcopenia and depressive symptoms, underscoring the importance of considering both conditions in comprehensive health management strategies for high-risk adults.
Our study provided new insights into the nature of the relationship between sarcopenia and depressive symptoms.To the best of our knowledge, there have been no studies elucidating the longitudinal reciprocal association between sarcopenia and depression.Similar to cross-sectional studies and meta-analysis, we also found a significant association between depression and sarcopenia [17,18].However, longitudinal research may offer more robust insights regarding the temporal direction of the observed associations.Recently, two observational studies reported a longitudinal effect of depression on sarcopenia onset [20,44] utilizing two-sample Mendelian randomization and reported evidence suggesting that depression symptoms could potentially result in a decrease in appendicular skeletal muscle mass but not vice versa [20].In contrast to previous studies, our study not only identified depressive symptoms as a precursor to sarcopenia but also uncovered a notably stronger temporal sequence effect of sarcopenia leading to depressive symptoms.
Importantly, this finding held true for both middle-aged and older adults.In terms of sex disparities, while males comprised a majority of the China community population due to socioeconomic and traditional cultural factors, females were more susceptible to developing sarcopenia and depressive symptoms, which is consistent with previous studies [44][45][46].
Diabetes, metabolic syndrome, cardiovascular diseases, and physical inactivity, all of which are associated with an increased risk of depression, have been linked to sarcopenia [47,48].Neurotrophins, chronic inflammation, and oxidative stress might play a role in the shared pathophysiological mechanisms bridging sarcopenia and depressive symptoms [49][50][51].Previous studies had shown that decreased muscle mass and strength have significant immune and redox effects by affecting the expression of neurotrophins (such as BDNF) and pro-inflammatory cytokines (CRP, IL-6, TNF-α), which consequently affect the development of depressive symptoms [51][52][53][54][55]. Another plausible explanation might be that the loss of skeletal muscle mass and function leads to less physical activity and malnutrition, which accordingly increases the risk of occurrence of depressive symptoms [27,56].
Increased evidence has demonstrated that a lack of physical activity and lower nutrition intake (e.g., protein, fruit, vegetables, and vitamin D) are associated with the upregulation of chronic inflammation and oxidative stress, which in turn lead to depressive symptoms and sarcopenia [57][58][59][60][61][62][63].To date, several feasible interventions used to improve sarcopenia have been proven effective in older adults, such as resistance training, protein supplement, and pharmacologic intervention [64][65][66][67].Given the absence of primary prevention guidelines for sarcopenia in China as of yet, implementing early depression screening in community-based care services and encouraging individuals to maintain a healthy lifestyle could aid in the identification of those at increased risk of developing sarcopenia [44].Assessing sarcopenia in male and older participants may facilitate the early detection and treatment of depressive symptoms and, accordingly, could reduce the incidence of major depressive symptoms disorder.While this study may have focused on sarcopenia and depression prevention in China, its implications extend to populations globally.By advocating for the adoption of healthy lifestyle practices and the implementation of early screening protocols for depression, this discovery enriches the broader discourse on proactive strategies aimed at mitigating sarcopenia, combating depression, and fostering healthy aging on a global scale.
While our study highlighted the stronger directional influence of early sarcopenia on subsequent depressive symptoms compared to the reverse pathway, we acknowledge the significance of addressing the impact of depressive symptoms on sarcopenia.Our findings suggest that early screening for depressive symptoms may also be beneficial in mitigating the occurrence of sarcopenia, particularly among females and middle-aged individuals.Given that both sarcopenia and depressive symptoms could independently predict cardiovascular disease and mortality, the early identification and intervention of these conditions may be helpful to prevent overall poor outcomes [10,30,68].
This research possesses several strengths.To begin with, individuals who were 45 years of age or older were selected as participants from the CHARLS study, a populationbased prospective cohort that is nationally representative.In this study, well-validated measures were used to assess depressive symptoms and sarcopenia status.Furthermore, to clarify the temporal bidirectional relationships between these two factors, our investigation employed a cross-lagged panel model, compared to other prospective studies that followed a unidirectional approach.Finally, it is important to note that this research is the first endeavor to investigate the possible temporal progression of alterations in depressive symptoms and sarcopenia in the Chinese population.As such, our results merit additional verification and investigation.
Potential limitations also need to be addressed.First, the research was an observational study, and recall bias in questionnaire surveys was unavoidable.Furthermore, it should be noted that although the statistical significance of the reciprocal associations between depressive symptoms and sarcopenia was preserved even after controlling for numerous potential confounding variables, there might be additional confounding factors that were not accounted for in our research.Additional potential confounding variables, including but not limited to dietary intake, body fat mass, and other psychiatric disorders like eating disorders and substance use disorder, as well as inflammation factors such as IL-6 and TNF-α, should be considered in future research.Third, it is important to highlight that the evaluation of gait speed was conducted at a distance of 2.5 m, which deviates from the traditional 6 m standard.However, an earlier investigation indicated that the gait speed measurements acquired by CHARLS were consistent with those obtained in China, suggesting that the distance discrepancy did not have an effect on the gait speed results [68].Fourth, the data utilized spanned from 2011 to 2018, potentially limiting its relevance to current circumstances.Factors such as the COVID-19 pandemic in 2020 could have influenced the prevalence of both sarcopenia and depressive symptoms, necessitating updated data to accurately depict the current state of conditions.Finally, a significant proportion of individuals lacked initial or subsequent assessments for depressive symptoms and sarcopenia; this incomplete follow-up may introduce potential bias into our findings, despite the substantial number of eligible participants involved in the study.

Conclusions
In conclusion, this study identified a longitudinal reciprocal association between sarcopenia and depressive symptoms.Individuals who are male and older than 60 are more sensitivity to sarcopenia and develop an increased risk of depressive symptoms in later life.Female and middle-aged adults are vulnerable to depressive symptoms and exhibited a higher likelihood to develop sarcopenia.Cross-lagged path coefficients suggest that the association between initial sarcopenia and subsequent depressive symptoms is more likely to be significant than the reverse pathway.The early assessment and improvement of sarcopenia and depressive symptoms may be beneficial to reduce the risk of poor overall health and foster the process of healthy aging within the Asian population.Further research is necessary to unravel this relationship and elucidate the underlying mechanism.Given the established association between sarcopenia and depressive symptoms, it is imperative to develop further cost-effective therapies.

Figure 1 .
Figure 1.Flow chart of the study population.CHARLS: China Health and Retirement Longitudinal Study; CES-D-10: 10-item Center for Epidemiologic Studies Depressive Scale.

Figure 1 .
Figure 1.Flow chart of the study population.CHARLS: China Health and Retirement Longitudinal Study; CES-D-10: 10-item Center for Epidemiologic Studies Depressive Scale.

Figure 2 .
Figure 2. Survival curves of cumulative probabilities of outcomes.(A) Overall population for depressive symptoms.(B) Middle-aged population for depressive symptoms.(C) Old population for depressive symptoms.(D) Overall population for sarcopenia.(E) Middle-aged population for sarcopenia.(F) Old population for sarcopenia.

Figure 2 .
Figure 2. Survival curves of cumulative probabilities of outcomes.(A) Overall population for depressive symptoms.(B) Middle-aged population for depressive symptoms.(C) Old population for depressive symptoms.(D) Overall population for sarcopenia.(E) Middle-aged population for sarcopenia.(F) Old population for sarcopenia.

Figure 3 .
Figure 3. Association between sarcopenia status and depressive symptoms by sex, age, sleep duration, and physical comorbidities.All HRs were calculated by adjusting age, sex, residential region, education level, average household income, BMI, marital status, smoking status, alcohol intake, physical comorbidities, sleep duration, and MMSE score.CI, confidence interval; HR, hazard ratio.

Figure 3 .
Figure 3. Association between sarcopenia status and depressive symptoms by sex, age, sleep duration, and physical comorbidities.All HRs were calculated by adjusting age, sex, residential region, education level, average household income, BMI, marital status, smoking status, alcohol intake, physical comorbidities, sleep duration, and MMSE score.CI, confidence interval; HR, hazard ratio.

Figure 3 .
Figure3.Association between sarcopenia status and depressive symptoms by sex, age, sleep duration, and physical comorbidities.All HRs were calculated by adjusting age, sex, residential region, education level, average household income, BMI, marital status, smoking status, alcohol intake, physical comorbidities, sleep duration, and MMSE score.CI, confidence interval; HR, hazard ratio.

Figure 4 .
Figure 4. Association between depressive symptoms status and incident sarcopenia by sex, sleep duration, and physical comorbidities.All HRs were calculated by adjusting age, sex, residential region, education level, average household income, BMI, marital status, smoking status, alcohol intake, physical comorbidities, sleep duration, and MMSE score.CI, confidence interval; HR, hazard ratio.

Figure 4 .
Figure 4. Association between depressive symptoms status and incident sarcopenia by sex, sleep duration, and physical comorbidities.All HRs were calculated by adjusting age, sex, residential region, education level, average household income, BMI, marital status, smoking status, alcohol intake, physical comorbidities, sleep duration, and MMSE score.CI, confidence interval; HR, hazard ratio.

Figure 5 .
Figure 5. Cross-lagged panel models applied to assess depression symptoms and sarcopenia status, and mediating effect of healthy lifestyle among Chinese adults.(A) In the overall population; (B) in the middle-aged population; (C) in the older-aged population.*** means p <0.001.

Figure 5 .
Figure 5. Cross-lagged panel models applied to assess depression symptoms and sarcopenia status, and mediating effect of healthy lifestyle among Chinese adults.(A) In the overall population; (B) in the middle-aged population; (C) in the older-aged population.** means p < 0.01, *** means p < 0.001.

Figure A1 .
Figure A1.Association between sarcopenia status and each single depressive symptom.

Figure A2 .
Figure A2.Association between sarcopenia indicators and depressive symptoms.Table A1.Inclusion and exclusion criteria for the two prospective cohorts.Cohort 1: Depressive Symptoms Cohort Cohort 2: Sarcopenia Cohort Inclusion Criteria Exclusion Criteria Inclusion Criteria Exclusion Criteria

Table 1 .
Characteristics of all participants at baseline by sarcopenia.

Table 1 .
Characteristics of all participants at baseline by sarcopenia.

Table 2 .
Risk of depression associated with baseline sarcopenia status.

Table 2 .
Risk of depression associated with baseline sarcopenia status.

Table 3 .
Characteristics of all participants at baseline by depressive symptoms.

Table 4 .
Risk of sarcopenia associated with baseline depression.

Table A1 .
Inclusion and exclusion criteria for the two prospective cohorts.

1: Depressive Symptoms Cohort Cohort 2: Sarcopenia Cohort Inclusion Criteria Exclusion Criteria Inclusion Criteria Exclusion Criteria
Individuals aged 45 years or older in 2011 Missing data on all depressive symptoms in 2013, 2015, or 2018 Individuals aged 45 years or older in 2011 Absence of data pertaining to all components associated with sarcopenia in the years 2013, 2015, or 2018

Table A1 .
Inclusion and exclusion criteria for the two prospective cohorts.

Table A2 .
Chronic diseases of all participants at baseline by sarcopenia.

Table A3 .
Chronic disease of all participants at baseline by depressive symptoms.

Table A4 .
Sensitivity analysis regarding risk of depressive symptoms with baseline sarcopenia status.