Abstract
Background
Both low muscle strength and depressive symptoms are common among middle-aged and older adults. However, their bilateral association has not been thoroughly investigated. This study examined whether muscle strength and depression are bilaterally associated using data from a prospective cohort in China.
Methods
A total of 6,961 participants without depression and 6,817 participants without low muscle strength from the China Health and Retirement Longitudinal Study were included to investigate the association of muscle strength with depression and of depressive symptoms with low muscle strength, respectively. The 10-item Center for Epidemiologic Studies Depression Scale (CESD-10) was used to assess depressive symptoms, with a cut-off score of 10. Handgrip strength and the time taken to complete the Five Times Sit-to-Stand Test (5TSTS) were used to evaluate muscle strength. Cox regression was applied to assess the associations after adjusting for relevant variables. Restricted cubic splines (RCS) were used to explore potential non-linear relationships.
Results
There was a significant association between handgrip strength (HR: 0.98; 95% CI: 0.98–0.99) and time of 5TSTS (HR: 1.02; 95% CI: 1.01–1.03) with the 7-year incident depressive symptoms. Similarly, depressive symptoms were significantly associated with the 4-year incident low muscle strength, as defined by both handgrip strength (HR: 1.04; 95% CI: 1.02–1.05) and time of 5TSTS (HR: 1.02; 95% CI: 1.01–1.04). No evidence of non-linear associations was detected.
Limitation
Depressive symptoms were assessed using the CESD-10 scale rather than a clinical diagnosis.
Conclusions
A bilateral longitudinal association exists between muscle strength and depression.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-025-24915-2.
Keywords: Muscle strength, Depression, Bilateral association, Cox regression, Prospective cohort
Introduction
Depressive symptoms are common among middle-aged and older adults. Globally, it is estimated that over 280 million people suffer from depression, contributing to approximately 47 million disability-adjusted life years in 2019 [1]. According to a recent cross-sectional survey, the prevalence of depressive symptoms among the general population in mainland China was 15.81% [2]. Low muscle strength is another widespread condition among older adults. Data from the National Health and Nutrition Examination Survey (NHANES) in the United States showed that the prevalence of reduced muscle strength among adults aged 60 and above was 18% [3]. A meta-analysis reported that the overall prevalence of sarcopenia (defined as low muscle mass combined with low muscle strength and/or low physical performance) among community-dwelling Chinese adults aged over 65 years was 17.4% [4]. Epidemiological studies have demonstrated that both depression and low muscle strength are associated with a range of adverse outcomes and can significantly impair quality of life [5, 6].
Given that depression and low muscle strength are both common and share several risk factors [7, 8], numerous studies—both cross-sectional and prospective—have explored the association between muscle strength and depression. For instance, one NHANES-based study found a significant association between grip strength and depression [9]. Additionally, two large-scale prospective cohort studies from China and the UK also confirmed a longitudinal relationship between muscle strength and the risk of depression [10, 11]. However, the influence of depressive symptoms on the incident low muscle strength has been scarcely investigated, and the potential bilateral association between the two conditions has not been clearly established. This study aims to explore the bilateral relationship between muscle strength and depression among middle-aged and older adults using data from a prospective cohort in China.
Methods
Source of data
Our study utilized data from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative, ongoing longitudinal cohort study in China. The survey began in 2011, with follow-up waves conducted in 2013, 2015, and 2018. It collects comprehensive information on respondents and their families, including demographic characteristics, intergenerational financial transfers, health status, medical care and insurance, employment, income, expenditures, and assets. In addition, CHARLS includes 13 physical measurements and the collection of blood samples. Further details about the study have been documented in previous research [12]. CHARLS was approved by the Ethics Review Committee of Peking University, and all participants provided informed consent prior to enrollment.
Assessment of muscle strength
In CHARLS, muscle strength was evaluated using both handgrip strength and lower limb muscle strength measurements in 2011, 2013, and 2015. Handgrip strength was assessed in both the dominant and non-dominant hands using a Yuejian™ WL-1000 dynamometer (Nantong, China). Participants were instructed to exert maximal force while squeezing the device, with two trials performed for each hand while maintaining a 90° elbow angle. For analysis, the average of the maximum values from each hand was used. If one hand could not be measured for any reason, the highest recorded value from the other hand was used. Low handgrip strength was defined based on the 2019 Asian Working Group for Sarcopenia (AWGS) criteria: less than 28 kg for males and less than 18 kg for females [13].
Lower limb muscle strength was assessed using the Five Times Sit-to-Stand Test (5TSTS). Participants were instructed to cross their arms over their chests and perform five consecutive sit-to-stand movements as quickly as possible, without pausing or using their arms for support. Investigators used stopwatches to objectively record the time taken to complete the task. According to the AWGS 2019 criteria, a completion time of 12 s or more was considered indicative of low lower limb muscle strength [13].
Assessment of depressive symptoms
Depressive symptoms were assessed in 2011, 2013, 2015, and 2018 using the 10-item Center for Epidemiologic Studies Depression Scale (CESD-10), a shortened version of the original 20-item CESD. Participants reported the frequency of 10 symptoms experienced over the past week, with response options categorized as: 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 scored from 0 to 3, with higher scores indicating greater symptom frequency. Total scores ranged from 0 to 30. Consistent with previous studies, a score of ≥ 20 was considered indicative of significant depressive symptoms, while a cut-off of ≥ 10 was also used to identify participants with depressive symptoms; scores < 10 defined the no-symptom group [14]. The CESD-10 has been widely validated for use in general populations and has demonstrated strong reliability and validity among community-dwelling older adults in China [15].
Assessment of covariates
Sociodemographic variables included age, gender, education level, marital status, number of children, and household income. Education was classified into three categories: elementary school or lower, middle school, and high school or higher. Marital status was categorized as married (living together) or unmarried (including married but living apart, separated, divorced, widowed, or never married). Behavioral covariates included smoking status (smoker or non-smoker), alcohol use (drinker or non-drinker), and weight loss. Health-related covariates included obesity (body mass index ≥ 30), aging-associated cognitive decline [16, 17], hypertension (self-reported, systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, or use of antihypertensive medications), diabetes (self-reported, hemoglobin A1c ≥ 6.5%, or use of antidiabetic medications), dyslipidemia (self-reported, total cholesterol ≥ 240 mg/dL, or use of lipid-lowering medications), and the self-reported number of comorbidities (excluding hypertension, diabetes, and dyslipidemia).
Statistical analysis
The baseline survey recruited 17,705 respondents, covering 150 counties of 28 provinces in China. To evaluate the association of muscle strength with depression, participants at baseline without information on muscle strength were excluded. Those without information on depressive symptoms or diagnosed with depression at baseline were also excluded. In addition, we excluded participants without follow-up information on depression. Similarly, to evaluate the association of depressive symptoms with low muscle strength, participants without information on depression at baseline were excluded. Those without information on muscle strength or diagnosed with low muscle strength at baseline were also excluded. In addition, we excluded participants without follow-up information on muscle strength. The flow chart of sample selection was presented in Fig. 1. Missing covariate data were handled using multiple imputation by chained equations to reduce potential bias in statistical inference.
Fig. 1.
The flow chart of sample selection
We analyzed data with R software (R Development Core Team, Version 4.1.3) In the descriptive statistical analyses, continuous variables were expressed as mean (standard deviation, SD) or median (interquartile ranges, IQR), and they were compared by the student’s t-test or Wilcoxon rank test. Categorical variables were expressed as numbers (percentages) and compared by the Chi-square test. P values less than 0.05 were considered of significance and Bonferroni correction was applied to adjust for multiple testing (0.05/2).
Cox proportional hazard models were employed to investigate the bilateral association between muscle strength and depression, using follow-up time as the time-scale variable. Results were expressed as hazard ratios (HRs) and 95% confidence intervals (CIs). Three different Cox models were fitted. Model 1 was not adjusted for any confounders. Model 2 was adjusted for age, gender, education, marital status, and income. Model 3 was adjusted for all the potential covariates. To assess potential non-linear associations, exposure variables were modeled using restricted cubic splines (RCS) within Cox models. The number of knots was selected based on the smallest Akaike Information Criterion. Non-linearity was tested using the Wald test for RCS coefficients.
We conducted several supplementary analyses to validate our findings. First, we applied a multivariable generalized linear mixed model with random intercepts and slopes to account for repeated measures and participant-specific trajectories. Second, we used the median follow-up time as the time-to-event in the Cox proportional hazards models. Third, we fitted an interval-censored survival model, given that the exact timing of events was unknown. Fourth, we performed subgroup analyses and tested interactions in vulnerable populations, including older adults (≥ 60 years) with multimorbidity (≥ 2 diseases), individuals with low income (< median household income), and those with fewer children (< 2 children).
Results
We first assessed the association between muscle strength and depression. Table 1 summarizes baseline characteristics by incident depressive symptoms. Among 6,961 participants (median age 57.0 years [IQR 50–64]; 57.8% male), most had an elementary school education and were married. Compared with participants without depressive symptoms, those with symptoms were younger, less often male, had lower education and income, fewer children, and were less likely to smoke or drink. They were more likely to exhibit mild cognitive decline and multimorbidity, had lower handgrip strength, and required more time to complete the 5TSTS.
Table 1.
Baseline characteristics of participants according to incident depressive symptoms
| Characteristics | Total | Non-depressive | Depressive | P-value |
|---|---|---|---|---|
| N | 6961 | 4254 | 2707 | |
| Age (years), median (IQR) | 57.0 (50.0–64.0) | 57.0 (50.0–64.0) | 56.0 (50.0–63.0) | 0.003 |
| Male, n (%) | 3671 (52.8) | 2460 (57.8) | 1211 (44.8) | < 0.001 |
| Education, n (%) | < 0.001 | |||
| Primary school | 4272 (61.4) | 2429 (57.1) | 1843 (68.1) | |
| Middle school | 1685 (24.2) | 1095 (25.7) | 590 (21.8) | |
| High school or above | 1004 (14.4) | 730 (17.2) | 274 (10.1) | |
| Married and lived together, n (%) | 6077 (87.3) | 3722 (87.5) | 2355 (87.0) | 0.568 |
| Number of children, median (IQR) | 2.0 (2.0–3.0) | 2.0 (2.0–3.0) | 2.0 (2.0–3.0) | < 0.001 |
| Income (yuan), median (IQR) | 31,200 (10400–55240) | 31,200 (12400–61950) | 26,508 (7975–46140) | < 0.001 |
| Obesity, n (%) | 368 (5.3) | 217 (5.1) | 151 (5.6) | 0.417 |
| Weight loss, n (%) | 376 (5.4) | 199 (4.7) | 177 (6.5) | < 0.001 |
| Smoke, n (%) | 2299 (33.0) | 1494 (35.1) | 805 (29.7) | < 0.001 |
| Alcohol use, n (%) | 2548 (36.6) | 1696 (39.9) | 3 852 ( 31.5) | < 0.001 |
| Hypertension, n (%) | 2669 (38.3) | 1644 (38.6) | 1025 (37.9) | 0.530 |
| Diabetes, n (%) | 1275 (18.3) | 809 (19.0) | 466 ( 17.2) | 0.062 |
| Dyslipidemia, n (%) | 1418 (20.4) | 504 (16.2) | 335 (16.4) | 0.395 |
| Mild cognitive decline, n (%) | 1094 (15.7) | 597 (14.0) | 497 ( 18.4) | < 0.001 |
| Number of comorbidities, median (IQR) | 0.0 (0.0–1.0) | 0.0 (0.0, 1.0) | 1.0 (0.0–1.0) | < 0.001 |
| Handgrip strength (Kg), median (IQR) | 34.0 (27.5–42.0) | 35.2 (29.0-43.2) | 32.0 (26.0–40.0) | < 0.001 |
| Time of 5TSTS (s), median (IQR) | 9.5 (7.8–11.7) | 9.3 (7.6–11.5) | 9.8 (8.0–12.0) | < 0.001 |
IQR interquartile range, 5TSTS Five Times Sit to Stand Test
Table 2 shows the association between muscle strength and incident depressive symptoms. After 7 years of follow-up, 2,707 participants (38.9%) had incident depressive symptoms. Handgrip strength and time of 5TSTS were both strongly associated with the incident depressive symptoms in Model 1, and these associations remained significant in Models 2 and 3 after adjusting for potential covariates. The association remained significant when handgrip strength and time of 5TSTS were grouped into quantiles. No non-linear relationship was detected between handgrip strength and the incidence of depression (P for non-linearity = 0.441; Fig. 2A). Similarly, the association between time of 5TSTS and the incident depressive symptoms was linear (P for non-linearity = 0.089; Fig. 2B).
Table 2.
Association of muscle strength with incident depressive symptoms
| No. of cases/ participants | Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | RD | HR (95% CI) | RD | HR (95% CI) | RD | ||
| Handgrip strength, Kg | 2707/6961 | 0.98 (0.97–0.98) | 0.98 (0.97–0.99) | 0.98 (0.98–0.99) | |||
| First quantile (< 27.5) | 833/1741 | Reference | Reference | Reference | |||
| Second quantile (27.5–34.0) | 736/1740 | 0.80 (0.72–0.88) | -0.08 | 0.82 (0.74–0.98) | -0.07 | 0.82 (0.74–0.91) | -0.07 |
| Third quantile (34.0–42.0) | 615/1740 | 0.63 (0.57–0.70) | -0.15 | 0.70 (0.62–0.79) | -0.13 | 0.71 (0.63–0.80) | -0.13 |
| Fourth quantile ( > = 42.0) | 523/1740 | 0.51 (0.45–0.56) | -0.22 | 0.59 (0.51–0.68) | -0.19 | 0.61 (0.53–0.71) | -0.18 |
| P for trend | < 0.001 | < 0.001 | < 0.001 | ||||
| Time of 5TSTS, seconds | 2707/6961 | 1.03 (1.02–1.03) | 1.02 (1.01–1.03) | 1.02 (1.01–1.03) | |||
| First quantile (< 7.8) | 591/1741 | Reference | Reference | Reference | |||
| Second quantile (7.8–9.5) | 661/1740 | 1.15 (1.03–1.29) | 0.04 | 1.10 (0.98–1.22) | 0.03 | 1.09 (0.97–1.21) | 0.03 |
| Third quantile (9.5–11.7) | 698/1740 | 1.27 (1.14–1.42) | 0.07 | 1.18 (1.05–1.31) | 0.06 | 1.17 (1.04–1.30) | 0.05 |
| Fourth quantile ( > = 11.7) | 757/1740 | 1.47 (1.32–1.64) | 0.12 | 1.32 (1.18–1.47) | 0.10 | 1.30 (1.16–1.45) | 0.09 |
| P for trend | < 0.001 | < 0.001 | < 0.001 | ||||
CI confidence interval, HR hazard ratio, RD risk difference, 5TSTS Five Times Sit to Stand Test; Model 1 was the basic model. Mode 2 was adjusted for age, gender, education, income, marriage; Model 3 was further adjusted for smoke, alcohol use, obesity, hypertension, diabetes, dyslipidemia, number of comorbidities, number of child, weight loss, and mild cognitive decline
Fig. 2.
The nonlinear association between muscle strength and depression was assessed using a fully adjusted Cox regression model. Restricted cubic splines with three knots were applied, with the median value set as the reference. A The non-linear association of handgrip strength with the incident depressive symptoms. B The non-linear association of time of 5TSTS with the incident depressive symptoms. C The non-linear association of CESD-10 score with the incident low muscle strength defined by handgrip strength. D The non-linear association of CESD-10 score with the incident low muscle strength defined by time of 5TSTS. HR: hazard ratio; CI: confidence interval; 5TSTS: Five Times Sit to Stand Test; CESD-10: The 10-item Center for Epidemiologic Studies Depression Scale
Next, we examined the association between depressive symptoms and low muscle strength. Table 3 presents participant characteristics according to muscle strength status, defined by handgrip strength and 5TSTS performance. A total of 6,817 participants were included. Compared with those with normal muscle strength, participants with low muscle strength were older, less educated, had lower income, were less likely to be married, had more children, and were more likely to have hypertension and multimorbidity. They also had higher CESD-10 scores.
Table 3.
Baseline characteristics of participants according to muscle strength
| Characteristics | Total | Normal muscle strength (handgrip strength) | Low muscle strength (handgrip strength) | P-value | Normal muscle strength (time of 5TSTS) | Low muscle strength (time of 5TSTS) | P-value |
|---|---|---|---|---|---|---|---|
| N | 6817 | 6187 | 630 | 5399 | 1418 | ||
| Age (Median, Years) | 56 (49.0–62.0) | 56.0 (49.0–62.0) | 62.0 (56.0–69.0) | < 0.001 | 55.0 (49.0–61.0) | 59.0 (53.0–66.0) | < 0.001 |
| Male (%) | 3378 (49.6) | 3053 (49.93) | 325 (51.6) | 0.303 | 2768 (51.3) | 610 (43.0) | < 0.001 |
| Education (%) | < 0.001 | < 0.001 | |||||
| Primary school | 4322 (63.4) | 3825 (61.8) | 492 (78.9) | 3232 (59.9) | 1090 (76.9) | ||
| Middle school | 1627 (23.9) | 1535 (24.8) | 92 (14.6) | 1405 (26.0) | 222 (15.7) | ||
| High school or above | 868 (12.7) | 827 (13.4) | 41 (6.5) | 762 (14.1) | 106 (7.5) | ||
| Income (Median, Yuan) | 19,880 (4100–32000) | 20,360 (4100–33080) | 11,860 (2510–26160) | < 0.001 | 20,800 (4100–33540) | 14,400 (3800–28975) | < 0.016 |
| Married and lived together | 5909 (86.7) | 5393 (87.2) | 516 (81.9) | < 0.001 | 4732 (87.6) | 1177 (83.0) | < 0.001 |
| Number of child, median (IQR) | 2.0 (2.0–3.0) | 2.0 (2.0–3.0) | 3.0 (2.0–4.0) | < 0.001 | 2.0 (2.0–3.0) | 3.0 (2.0–4.0) | < 0.001 |
| Obesity, n (%) | 328 (4.8) | 305(4.9) | 23 (3.7) | 0.183 | 245 (4.5) | 83 (5.9)) | 0.047 |
| Weight loss, n (%) | 474 (7.0) | 408 (6.6) | 66 (10.5) | < 0.001 | 369 (6.8) | 105 (7.4) | 0.489 |
| Smoke (%) | 2202 (32.3) | 1998 (32.3) | 204 (32.4) | 0.999 | 1791 (33.2) | 411 (29.0) | 0.003 |
| Alcohol use (%) | 2454 (36.0) | 2230 (36.0) | 224 (35.6) | 0.842 | 2029 (37.6) | 425 (30.0) | < 0.001 |
| Hypertension (%) | 2433 (35.7) | 2173 (35.1) | 260 (41.3) | 0.002 | 1837 (34.0) | 596 (42.0) | < 0.001 |
| Diabetes (%) | 1180 (17.3) | 1065 (17.2) | 115 (18.3) | 0.547 | 937 (17.4) | 243 (17.1) | 0.878 |
| Dyslipidemia (%) | 1414 (20.7) | 1274 (20.6) | 140 (22.2) | 0.363 | 1148 (21.3) | 266 (18.8) | 0.042 |
| Mild cognitive decline | 1160 (17.0) | 1042 (16.8) | 118 (18.7) | 0.252 | 906 (16.8) | 254 (17.9) | 0.332 |
| Number of Comorbidities | 1.0 (0.0–1.0) | 1.0 (0.0–1.0) | 1.0 (0.0–2.0) | < 0.001 | 1.0 (0.0–1.0) | 1.0 (0.0–2.0) | < 0.001 |
| Score of CESD-10 | 6.0 (3.0–11.0) | 6.0 (3.0–11.0) | 8.0 (4.0–14.0) | < 0.001 | 6.0 (3.0–10.0) | 7.0 (4.0–12.0) | < 0.001 |
IQR interquartile range, CESD-10 10-item Center for Epidemiologic Studies Depression Scale, 5TSTS Five Times Sit to Stand Test
Table 4 shows the association between depressive symptoms and incident low muscle strength. After 4 years of follow-up, 630 and 1,418 participants were identified as having low muscle strength based on handgrip strength and time of 5TSTS, respectively. Higher CESD-10 scores were strongly associated with low muscle strength (as defined by handgrip strength) in Model 1, and the association remained significant in Models 2 and 3 after adjusting for covariates. The association also remained significant when CESD-10 scores were grouped into quantiles. No non-linear relationship was observed between CESD-10 score and low muscle strength defined by handgrip strength (P for non-linearity = 0.169; Fig. 2C). Similar results were observed for the association between CESD-10 score and low muscle strength defined by time of 5TSTS (P for non-linearity = 0.649; Fig. 2D).
Table 4.
Association of depressive symptoms with low muscle strength
| Incident low muscle strength (handgrip strength) | |||||||
|---|---|---|---|---|---|---|---|
| No. of cases/ participants | Model 1 | Model 2 | Model 3 | ||||
| HR (95% CI) | RD | HR (95% CI) | RD | HR (95% CI) | RD | ||
| CESD-10 score | 630/6817 | 1.05 (1.04-1.06) | 1.04 (1.03-1.06) | 1.04 (1.02-1.05) | |||
| First quantile (0-3) | 94/1365 | Reference | Reference | Reference | |||
| Second quantile (3-7) | 123/1656 | 1.08 (0.85-1.38) | 0.01 | 1.02 (0.80-1.30) | 0e+00 | 0.99 (0.78-1.26) | 0e+00 |
| Third quantile (7-12) | 172/1975 | 1.14 (0.90-1.45) | 0.01 | 1.09 (0.86-1.39) | 1e-04 | 1.03 (0.81-1.32) | 0e+00 |
| Fourth quantile (12-30) | 241/1821 | 1.80 (1.45-2.24) | 0.06 | 1.58 (1.27-1.98) | 7e-04 | 1.42 (1.12-1.79) | 5e-04 |
| P for trend | <0.001 | <0.001 | <0.001 | ||||
| Incident low muscle strength (time of 5TSTS) | |||||||
|---|---|---|---|---|---|---|---|
| No. of cases/ participants | Model 1 | Model 2 | Model 3 | ||||
| HR (95% CI) | RD | HR (95% CI) | RD | HR (95% CI) | RD | ||
| CESD-10 score | 1418/6817 | 1.04 (1.03-1.04) | 1.02 (1.01-1.03) | 1.02 (1.01-1.03) | |||
| First quantile (0-3) | 227/1365 | Reference | Reference | Reference | |||
| Second quantile (3-7) | 299/1656 | 1.19 (1.01-1.39) | 0.03 | 1.13 (0.97-1.33) | 3e-03 | 1.111 (0.94-1.29) | 3e-03 |
| Third quantile (7-12) | 419/1975 | 1.30 (1.11-1.52) | 0.04 | 1.21 (1.03-1.41) | 5e-03 | 1.16 (0.99-1.36) | 4e-03 |
| Fourth quantile (12-30) | 473/1821 | 1.66 (1.43-1.93) | 0.10 | 1.38 (1.19-1.61) | 9e-03 | 1.29 (1.10-1.51) | 7e-03 |
| P for trend | <0.001 | <0.001 | 0.001 | ||||
CI confidence interval, HR hazard ratio, RD risk difference, 5TSTS Five Times Sit to Stand Test; Model 1 was the basic model. Mode 2 was adjusted for age, gender, education, income, marriage; Model 3 was further adjusted for smoke, alcohol use, obesity, hypertension, diabetes, dyslipidemia, number of comorbidities, number of child, weight loss, and mild cognitive decline
The results of the complementary analyses are presented in the Supplemental Tables. Overall, the generalized linear mixed model (Supplemental Table 1), the Cox regression model using the median follow-up time as the time-to-event (Supplemental Table 2), and the interval-censored survival model (Supplemental Table 3) yielded results consistent with the main analysis. In addition, subgroup analyses and interaction tests produced similar findings across different subgroups (Supplemental Table 4).
Discussion
In this study, we report, for the first time, a significant bidirectional longitudinal association between muscle strength and depression in a prospective cohort. This relationship remained robust after adjustment for relevant covariates and across multiple statistical models. Subgroup analyses indicated that the association was consistent across vulnerable populations, and RCS analyses confirmed its linearity.
Several prospective studies have reported associations between muscle strength and depressive symptoms. For instance, lower handgrip strength has been linked to the development of depressive symptoms in adults in cohorts from England (≥ 60 years) [18], Italy (≥ 65 years) [19], and China (≥ 45 years) [20]. Beyond that, there is evidence linking physical functioning with the trajectory of depressive symptoms, and vice versa. Brett et al. (2022) conducted a 19-year study of former NFL players, showing that declines in physical function predicted a steeper increase in depressive symptoms over time, independent of concussion history [21]. Similarly, Lever-van Milligen et al. assessed six-year trajectories of handgrip strength and physical function in individuals with depressive and/or anxiety disorders, finding that those with current depression and anxiety experienced greater declines in physical function compared with healthy controls [22].
Studies exploring the bidirectional relationship between muscle strength and depression remain limited, although low muscle strength is frequently observed as a clinical feature in individuals with depression. Bertoni et al. used SHARE (Survey on Health, Ageing and Retirement in Europe) panel data to investigate whether the onset of low muscle strength at a 4-year follow-up could be explained by pre-existing depressive symptoms (either at baseline or at the 2-year follow-up), or vice versa [23]. Using multivariate logistic regression, they found that the onset of depressive symptoms increased the risk of low muscle strength. However, low handgrip strength at baseline was not associated with the risk of incident depression during follow-up. This insignificant association may have occurred due to the relatively short follow-up period or the use of only a single indicator of muscle strength.
The mechanisms by which muscle strength affects depression remain unclear, but several hypotheses involving neurophysiological and psychosocial pathways have been proposed. First, individuals with depression have consistently been reported to exhibit functional and structural brain alterations, with the hippocampus—an area involved in emotional processing and stress regulation—being among the most frequently affected regions [24, 25]. Levels of brain-derived neurotrophic factor (BDNF), a key marker of neuronal growth and plasticity, are reduced in individuals with depression [26]. In contrast, greater muscle strength has been positively associated with larger hippocampal volume, improved white matter integrity, and enhanced neurocognitive performance. Notably, resistance training—which is known to elevate BDNF levels—may play a protective role against depressive symptoms [27, 28]. Second, chronic low-grade inflammation may contribute to the development of depression [8, 29]. In response to muscle contractions, skeletal muscle tissue releases various cytokines and myokines into the bloodstream, some of which possess anti-inflammatory properties that may help protect against mood disorders [30].
Several mechanisms may explain the effect of muscle strength on depression. First, depressive symptoms are often linked to social isolation and loneliness, both of which are associated with poor physical performance and disability—conditions closely related to low muscle strength [31]. Second, malnutrition, which is common among older adults with depression [32], is a well-established risk factor for reduced muscle strength in the elderly [33, 34]. Third, depression is characterized by symptoms such as diminished interest, weight changes, and fatigue—all of which may contribute to the development of low muscle strength. Additionally, overlapping biological and physiological pathways, such as inflammation, may further clarify the link between depressive symptoms and reduced muscle strength [8, 29, 30].
This study has the advantage of using large-scale longitudinal data to explore the possible bilateral association between muscle strength and depression. It expands upon the findings of previous research by providing evidence of a two-way relationship between the two. In addition, the study used both handgrip strength and chair stand performance to represent upper and lower limb strength, respectively, which enhances the reliability of the findings. While our findings indicate a prospective association between muscle strength and incident depression, this does not necessarily imply a causal relationship. The association may reflect shared underlying biological or psychosocial mechanisms—such as chronic inflammation, neuroendocrine dysregulation, or socioeconomic and lifestyle factors—rather than a direct effect of muscle strength on depression. Accordingly, our findings should be interpreted with caution.
Several limitations of this study should be acknowledged. First, depression was assessed using the CESD-10, a screening instrument rather than a gold-standard diagnostic interview. Although the CESD-10 has been validated in Chinese and other populations [35, 36], it is not without limitations, and outcome misclassification is possible. In the CHARLS prospective cohort, the CESD-10 was administered blindly with respect to future outcome ascertainment and applied uniformly across exposure groups; therefore, any misclassification is most plausibly non-differential, as sensitivity and specificity are unlikely to vary systematically by exposure status. Under such circumstances, hazard ratio estimates are generally biased toward the null, suggesting that the true associations may be stronger than those reported [37]. Second, certain important covariates—such as physical activity level and specific medication use (e.g., antidepressants)—were unavailable. These factors could influence both exposure and outcome, introducing residual confounding that limits causal interpretation. Although our associations remained robust after adjusting for available covariates, future studies should incorporate more detailed assessments of these variables. Sensitivity analyses using E-values, which estimate the minimum strength of association that an unmeasured confounder would need to nullify the observed results [38], may further strengthen causal inference. Third, although the longitudinal design supports temporal sequencing, reverse causation cannot be entirely excluded. Subclinical depressive symptoms or early declines in muscle strength may have been present at baseline, potentially influencing both exposure and outcome and biasing the observed relationships. Finally, while the CHARLS cohort is nationally representative of community-dwelling middle-aged and older adults in China, caution is warranted when generalizing these findings to other populations. Cultural norms, institutional contexts, and healthcare systems may influence both muscle strength and depression risk. Replication in diverse settings is needed to assess external validity and to explore potential effect modification by societal and environmental factors.
In conclusion, this study demonstrates a bidirectional longitudinal association between muscle strength and depression, underscoring a self-reinforcing cycle that may accelerate functional and psychological decline in older adults. Breaking this cycle will require integrated strategies that simultaneously target physical and mental health, particularly within primary care and community health systems where most older adults receive ongoing care. Combined exercise programs represent a promising dual-target intervention, with meta-analytic evidence showing benefits for muscle strength, depressive symptoms, and aspects of quality of life [39–41]. Defining the optimal modality, intensity, and duration of such interventions—and embedding them into routine care pathways—should be a priority for future research and policy.
Supplementary Information
Acknowledgements
We would like to express our sincere gratitude to the staff of the China Health and Retirement Longitudinal Study for providing the data.
Authors' contributions
Tong Guo designed the study and applied the data. Fanchao Meng and Shixiang Liang conducted the statistical analysis. Guangwei Zhao prepared the draft. Jun Liu and Xue Wang finished the manuscript and prepared the tables and figures. Ling Zhang and Gang Wang reviewed the data, statistical process, and the manuscript. All authors contributed to and have approved the final manuscript.
Funding
The study was supported by Beijing Anding Hospital Science Foundation (No. 12-2024), 2023 Sprout Science and Technology Innovation Fund by Beijing Anding Hospital, Capital Medical University (No. 13-2024), and STI2030-Major Projects (2021ZD0200600).
Data availability
The data are from CHARLS, which is openly available at https://charls.pku.edu.cn/en/.
Declarations
Ethics approval and consent to participate
CHARLS was approved by the Ethics Review Committee of Peking University, and all participants provided informed consent prior to enrollment.
Consent for publication
Not applicable. There is no individual level data in our publication.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Tong Guo and Fanchao Meng contributed equally to this work.
Contributor Information
Tong Guo, Email: guotong@mail.ccmu.edu.cn.
Gang Wang, Email: gangwangdoc@ccmu.edu.cn.
References
- 1.Vos T, Lim SS, Abbafati C, et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020;396(10258):1204–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Liu J, Wu J, Wang J, et al. Prevalence and associated factors for depressive symptoms among the general population from 31 provinces in china: the utility of social determinants of health theory. J Affect Disord. 2024;347:269–77. [DOI] [PubMed] [Google Scholar]
- 3.Looker AC, Wang C-Y. Prevalence of reduced muscle strength in older US adults: United States, 2011–2012 NCHS Data Brief. 2015;179:1–8. [PubMed]
- 4.Ren X, Zhang X, He Q, et al. Prevalence of sarcopenia in Chinese community-dwelling elderly: a systematic review. BMC Public Health. 2022;22(1):1702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Yang S, Li T, Yang H, et al. Association between muscle strength and health-related quality of life in a Chinese rural elderly population: a cross-sectional study. BMJ Open. 2020;10(1):e026560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sivertsen H, Bjørkløf GH, Engedal K, et al. Depression and quality of life in older persons: a review. Dement Geriatr Cogn Disord. 2015;40(5–6):311–39. [DOI] [PubMed] [Google Scholar]
- 7.Bano G, Trevisan C, Carraro S, et al. Inflammation and sarcopenia: a systematic review and meta-analysis. Maturitas. 2017;96:10–5. [DOI] [PubMed] [Google Scholar]
- 8.Miller AH, Raison CL. The role of inflammation in depression: from evolutionary imperative to modern treatment target. Nat Rev Immunol. 2016;16(1):22–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Qiu L, Ren Y, Li J, et al. Associations of muscle mass and strength with depression among US adults: A cross-sectional NHANES study. J Affect Disord. 2024;363:373–80. [DOI] [PubMed] [Google Scholar]
- 10.Bao M, Chao J, Sheng M, et al. Longitudinal association between muscle strength and depression in middle-aged and older adults: A 7-year prospective cohort study in China. J Affect Disord. 2022;301:81–6. [DOI] [PubMed] [Google Scholar]
- 11.Cabanas-Sánchez V, Esteban‐Cornejo I, Parra‐Soto S, et al. Muscle strength and incidence of depression and anxiety: findings from the UK biobank prospective cohort study. J Cachexia Sarcopenia Muscle. 2022;13(4):1983–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhao Y, Hu Y, Smith JP, et al. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol. 2014;43(1):61–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Chen L-K, Woo J, Assantachai P, et al. Asian working group for sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc. 2020;21(3):300–07. e2. [DOI] [PubMed] [Google Scholar]
- 14.Andresen EM, Malmgren JA, Carter WB, et al. Screening for depression in well older adults: evaluation of. Prev Med. 1994;10:77–84. [PubMed] [Google Scholar]
- 15.Chen H, Mui AC. Factorial validity of the center for epidemiologic studies depression scale short form in older population in China. Int Psychogeriatr. 2014;26(1):49–57. [DOI] [PubMed] [Google Scholar]
- 16.Levy R. Aging-associated cognitive decline. Int Psychogeriatr. 1994;6(1):63–8. [PubMed] [Google Scholar]
- 17.Richards M, Touchon J, Ledesert B, et al. Cognitive decline in ageing: are AAMI and AACD distinct entities? Int J Geriatr Psychiatry. 1999;14(7):534–40. [DOI] [PubMed] [Google Scholar]
- 18.Hamer M, Batty GD, Kivimaki M. Sarcopenic obesity and risk of new onset depressive symptoms in older adults: english longitudinal study of ageing. Int J Obes. 2015;39(12):1717–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Veronese N, Stubbs B, Trevisan C, et al. Poor physical performance predicts future onset of depression in elderly people: progetto Veneto anziani longitudinal study. Phys Ther. 2017;97(6):659–68. [DOI] [PubMed] [Google Scholar]
- 20.Zhao Z, Ji C, Liu Y, et al. Higher handgrip strength predicts a lower risk of depressive symptoms in rural Chinese populations. J Affect Disord. 2020;269:12–7. [DOI] [PubMed] [Google Scholar]
- 21.Brett BL, Kerr ZY, Walton SR, et al. Longitudinal trajectory of depression symptom severity and the influence of concussion history and physical function over a 19-year period among former National football league (NFL) players: an NFL-LONG study. J Neurol Neurosurg Psychiatry. 2022;93(3):272–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lever-van Milligen BA, Lamers F, Smit JH, et al. Six‐year trajectory of objective physical function in persons with depressive and anxiety disorders. Depress Anxiety. 2017;34(2):188–97. [DOI] [PubMed] [Google Scholar]
- 23.Bertoni M, Maggi S, Manzato E, et al. Depressive symptoms and muscle weakness: A two-way relation? Exp Gerontol. 2018;108:87–91. [DOI] [PubMed] [Google Scholar]
- 24.Kandola A, Ashdown-Franks G, Hendrikse J, et al. Physical activity and depression: towards Understanding the antidepressant mechanisms of physical activity. Neurosci Biobehav Rev. 2019;107:525–39. [DOI] [PubMed] [Google Scholar]
- 25.Moon C-M, Jeong G-W. Functional and morphological alterations associated with working memory dysfunction in patients with generalized anxiety disorder. Acta Radiol. 2017;58(3):344–52. [DOI] [PubMed] [Google Scholar]
- 26.Brunoni AR, Lopes M, Fregni F. A systematic review and meta-analysis of clinical studies on major depression and BDNF levels: implications for the role of neuroplasticity in depression. Int J Neuropsychopharmacol. 2008;11(8):1169–80. [DOI] [PubMed] [Google Scholar]
- 27.Firth JA, Smith L, Sarris J, et al. Handgrip strength is associated with hippocampal volume and white matter hyperintensities in major depression and healthy controls: a UK biobank study. Biopsychosoc Sci Med. 2020;82(1):39–46. [DOI] [PubMed] [Google Scholar]
- 28.Schuch FB, Stubbs B. The role of exercise in preventing and treating depression. Curr Sports Med Rep. 2019;18(8):299–304. [DOI] [PubMed] [Google Scholar]
- 29.Barnes J, Mondelli V, Pariante CM. Genetic contributions of inflammation to depression. Neuropsychopharmacology. 2017;42(1):81–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Köhler CA, Freitas TH, Maes Md, et al. Peripheral cytokine and chemokine alterations in depression: a meta-analysis of 82 studies. Acta Psychiatrica Scandinavica. 2017;135(5):373–87. [DOI] [PubMed] [Google Scholar]
- 31.Coyle CE, Dugan E. Social isolation, loneliness and health among older adults. J Aging Health. 2012;24(8):1346–63. [DOI] [PubMed] [Google Scholar]
- 32.Velázquez-Alva MC, Irigoyen-Camacho ME, Cabrer-Rosales MF, et al. Prevalence of malnutrition and depression in older adults living in nursing homes in Mexico City. Nutrients. 2020;12(8):2429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Singh DK, Manaf ZA, Yusoff NAM, et al. Correlation between nutritional status and comprehensive physical performance measures among older adults with undernourishment in residential institutions. Clin Interv Aging. 2014;9:1415–23. [DOI] [PMC free article] [PubMed]
- 34.Tramontano A, Veronese N, Giantin V, et al. Nutritional status, physical performance and disability in the elderly of the Peruvian Andes. Aging Clin Exp Res. 2016;28:1195–201. [DOI] [PubMed] [Google Scholar]
- 35.Boey KW. Cross-validation of a short form of the CES‐D in Chinese elderly. Int J Geriatr Psychiatry. 1999;14(8):608–17. [DOI] [PubMed] [Google Scholar]
- 36.Fu H, Si L, Guo R. What is the optimal cut-off point of the 10-item center for epidemiologic studies depression scale for screening depression among Chinese individuals aged 45 and over? An exploration using latent profile analysis. Front Psychiatry. 2022;13:820777. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yland JJ, Wesselink AK, Lash TL, et al. Misconceptions about the direction of bias from Nondifferential misclassification. Am J Epidemiol. 2022;191(8):1485–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.VanderWeele TJ, Ding P. Sensitivity analysis in observational research: introducing the E-value. Ann Intern Med. 2017;167(4):268–74. [DOI] [PubMed] [Google Scholar]
- 39.Noetel M, Sanders T, Gallardo-Gómez D, et al. Effect of exercise for depression: systematic review and network meta-analysis of randomised controlled trials. BMJ. 2024;384:e075847. [DOI] [PMC free article] [PubMed]
- 40.Zhang W, Liu X, Liu H, et al. Effects of aerobic and combined aerobic-resistance exercise on motor function in sedentary older adults: A randomized clinical trial. J Back Musculoskelet Rehabil. 2024;37(1):25–36. [DOI] [PubMed] [Google Scholar]
- 41.Mahmoudi A, Amirshaghaghi F, Aminzadeh R, et al. Effect of aerobic, resistance, and combined exercise training on depressive symptoms, quality of life, and muscle strength in healthy older adults: a systematic review and meta-analysis of randomized controlled trials. Biol Res Nurs. 2022;24(4):541–59. [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 from CHARLS, which is openly available at https://charls.pku.edu.cn/en/.