Longitudinal association of sleep duration with possible sarcopenia: evidence from CHARLS

Abstract

Objectives

There are limited data on the relationship between sleep duration and possible sarcopenia. Hence, this study aimed to investigate the associations of sleep duration with possible sarcopenia and its defining components based on the China Health and Retirement Longitudinal Study (CHARLS).

Design

A retrospective cohort study.

Setting

This study was conducted on participants aged over 45 years applying the 2011 baseline and 2015 follow-up survey from CHARLS covering 450 villages, 150 counties and 28 provinces.

Participants

Data from 5036 individuals (2568 men and 2468 women) free of possible sarcopenia at baseline were analysed.

Primary and secondary outcome measures

The dose-response relationship between sleep duration and possible sarcopenia.

Results

During 4 years of follow-up, 964 (19.14%) participants developed possible sarcopenia. Compared with participants who slept 6–8 hours per night, those with shorter sleep duration (<6 hours per night) were independently associated with 22% (OR, 1.22; 95% CI, 1.04 to 1.44) increased risk of developing possible sarcopenia and 27% (OR, 1.27; 95% CI, 1.04 to 1.57) increased risk of developing low handgrip strength after controlling for potential confounders. Long sleep duration (>8 hours per night) was not significantly associated with incident possible sarcopenia. The plots of restricted cubic splines exhibited an atypical inverse J-shaped association between sleep duration and possible sarcopenia. Subgroup analysis showed a stronger association between sleep duration and possible sarcopenia in participants aged 45–59 years and composed of male populations.

Conclusions

Short sleep duration was a potential risk factor for possible sarcopenia and low handgrip strength. The improvement of sleep duration should be considered a target in early preventive and administrative strategies against the development of handgrip strength decline and further reduced the occurrence of sarcopenia.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The China Health and Retirement Longitudinal Study is an ongoing, nationally representative longitudinal project among more than 17 000 individuals in China.

• This study is based on longitudinal analyses, and dose-response relationships are appraised.

• There is the possibility of some unmeasured confounders that are not adequately considered in this study due to data limitations.

Introduction

Sarcopenia is estimated to influence 10%–16% of older adults worldwide.¹ Emerging evidence suggests that sarcopenia increases the risk of multiple adverse outcomes in older persons, including fall, fracture, frailty, cognitive impairment, metabolic diseases, hospitalisation and mortality.^2–4 It is of great importance to explore the modifiable risk factors of sarcopenia, in order to provide early identification and timely prevention. Asian Working Group for Sarcopenia (AWGS) 2019 criteria introduced possible sarcopenia, defined by either low muscle strength or low physical performance, to promote early identification of individuals at risk of sarcopenia and raise awareness of sarcopenia prevention in primary care settings.⁵ Early identification of sarcopenia can aid in interventions before disease progression. Timely intervention for possible sarcopenia can prevent the further development of sarcopenia and ultimately improve the quality of life and reduce mortality among older adults.

There are many sociodemographic, behavioural and disease-related associated factors that contribute to the occurrence and development of sarcopenia in older adults, including sleep duration.^{5 6} Sleep plays an important role in the functioning of cells, organs and systems. Appropriate sleep can create a virtuous cycle that promotes human health. Studies have shown that both short and long sleep durations were associated with an increased risk of mortality.^7–9 Insufficient sleep can reduce the level of growth hormone (GH) and increase the level of cortisol and proinflammatory changes, which in turn increase the risk of cardiovascular disease, insulin resistance, diabetes and obesity.¹⁰

Current evidence on the association between sleep duration and the incidence of sarcopenia is inconsistent. For example, several studies have exhibited a U-shaped association between sleep duration and sarcopenia and indicated that both short and long sleep durations were potential risk factors for sarcopenia,^{6 11 12} while other studies suggested that only short or long sleep duration had a positive association with sarcopenia.^13–15 In contrast, studies from China showed that no significant associations were observed between sleep duration and sarcopenia.^{16 17} The inconsistent findings may potentially be attributed to differences in study design, measurement methods, study populations, sleep duration grouping and adjustments for covariates. Therefore, this inconsistency indicates the need for further research. In addition, to date, there have been no studies focusing on the association between sleep duration and possible sarcopenia.

Therefore, this study focuses on Chinese middle-aged and older adults and uses a large nationally representative longitudinal dataset from China Health and Retirement Longitudinal Study (CHARLS) to examine the association between sleep duration and possible sarcopenia. The aim is to provide objective scientific evidence for early prevention and intervention strategies for sarcopenia.

Methods

Study population

The present study is based on CHARLS, which is an ongoing, nationally representative longitudinal project among residents aged over 45 years covering 150 county-level units, 450 village-level units and more than 17 000 individuals in about 10 000 households at baseline survey. The baseline survey was conducted in 2011, and wave 2, wave 3 and wave 4 were followed in 2013, 2015 and 2018 respectively. The survey contents included demographic backgrounds, health status and functioning, social and economic status, and retirement information. The details of the study design and sampling method of CHARLS had been previously described by Zhao et al.¹⁸

Because possible sarcopenia was not measured in wave 4, the longitudinal data from 2011 to 2015 were used. Participants were excluded based on the following exclusion criteria: (1) aged <45 years; (2) missing information on sleep duration, possible sarcopenia or important covariates; (3) diagnosed sarcopenia; (4) diagnosed possible sarcopenia at the baseline; (5) health conditions such as memory-related diseases or psychiatric problems; (6) unable to perform tasks such as the five-time chair stand test; and (7) unable to communicate in Chinese. The CHARLS survey was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-11015), and all participants were required to sign informed consent.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Assessment of sleep duration

Sleep duration was available from the lifestyle and health behaviours section of the CHARLS questionnaire. More specifically, the item was ‘During the past month, how many hours of actual sleep did you get at night (average hours for one night)’. The participants provided the answers in hours and minutes. The actual time was categorised into three groups: <6 hours, 6–8 hours and >8 hours for one night, representing short, medium and long sleep duration.

Assessment of possible sarcopenia

Possible sarcopenia was assessed based on the criteria recommended by the AWGS 2019, including low muscle strength or low physical performance. Handgrip strength was measured in kilograms using a YuejianTM WL-1000 dynamometer. Each participant was tested twice for both hands. The average of the maximum values for both hands was recorded. The maximum values were taken for one hand if another hand could not complete for some reason. Low handgrip strength was defined as <28 kg for men and <18 kg for women. The five-time chair stand test was conducted to assess the physical performance. Each participant was asked to rise continuously five times keeping their arms folded across their chest from the height of the 47 cm chair. The total time spent was recorded. Low physical performance was defined as time ≥12 s for both men and women.

Potential confounders

According to prior knowledge, several potential confounders were considered to adjust the models, including demographic characteristics, lifestyle factors and health status. Specifically, demographic characteristics included age (45–60 years and ≥60 years), gender (male or female), marital status (married or other), educational level (primary school or below, middle school or above) and residence type (rural or urban). Lifestyle factors included smoking and drinking. Smoking was defined as >100 cigarettes in a lifetime.¹⁹ Drinking was defined as ever drunk alcoholic beverages, such as beer, wine or liquor, in the past year.²⁰ Inquiries addressed 12 health conditions through self-reported diagnosis, including hypertension, dyslipidaemia, diabetes, cancer, chronic lung diseases, liver diseases, heart diseases, stroke, kidney disease, stomach diseases, asthma and depression. Depression was defined by a Center for Epidemiological Studies Depression Scale (CES-D) score ≥10.²¹ The number of chronic diseases was classified into three groups: 0, 1 and ≥2. Body mass index (BMI, kg/m²) was calculated as the weight (kg) divided by the square of height (m²), which was categorised as <18.5, 18.5–23.9 and ≥24.0. The weight was measured by OmronTM HN-286 scale and the height was measured by SecaTM213 height metre.²²

Statistical analysis

Qualitative data were expressed as relative numbers and proportions, and differences between normal control and possible sarcopenia groups were compared using χ² test. Longitudinal association of sleep duration with possible sarcopenia was assessed using the logistic regression model. OR and 95% CI were used to quantitatively estimate the effect size. Sleep duration of 6–8 hours per night was set as a reference group. The potential confounders were gradually added into the models. For model 1, no confounders were included. For model 2, demographic characteristic factors were included. For model 3, lifestyle factors were added to model 2. Model 4 additionally included BMI and the number of chronic diseases. The exposure-response relation between sleep duration and possible sarcopenia was investigated using the restricted cubic spline with three knots. The associations between sleep duration and the components of possible sarcopenia were also analysed, namely low handgrip strength and physical performance. Subgroup analyses by gender and age were also performed. All the statistical analyses were conducted using R software, V.4.1.0 (University of Auckland, New Zealand). P<0.05 was considered statistically significant.

Results

Basic characteristics of the study population

Finally, a total of 5036 participants were enrolled. The detailed selection process is presented in figure 1. In this longitudinal study, 964 participants had developed possible sarcopenia from 2011 to 2015, with a cumulative incidence of 19.14%. The cumulative incidence of possible sarcopenia in the <6-hour, 6–8-hour and >8-hour groups was 23.83% (316/1326), 17.20% (570/3314) and 19.70% (78/396), respectively, with statistically significant difference (χ² =26.99, p<0.001). Table 1 showed the comparison of basic characteristics between participants with possible sarcopenia and normal controls. Compared with normal controls, participants with possible sarcopenia were found to be statistically older (≥60 years), more likely to be female, unmarried and residing in rural areas. Additionally, they exhibited a significantly lower level of education (primary school or below), lower BMI (<18.5 kg/m²) and a higher prevalence of chronic diseases.

Figure 1.

The flowchart of sampling selection process.

Table 1.

Comparison of baseline characteristics between possible sarcopenia and normal population (n=5036)

Variables	Total	Possible sarcopenia	Normal control	χ2	P value
Gender				6.31	0.012
Male	2568	456 (17.76)	2112 (82.24)
Female	2468	508 (20.58)	1960 (79.42)
Age, years				202.62	<0.001
45–59	3349	453 (13.53)	2896 (86.47)
≥60	1687	511 (30.29)	1176 (69.71)
Education				117.80	<0.001
Primary school or below	3172	754 (23.77)	2418 (76.23)
Middle school	1864	210 (11.27)	1654 (88.73)
Marital status				26.25	<0.001
Married	4381	790 (18.03)	3591 (81.97)
Other	655	174 (26.56)	481 (73.44)
Area of residence				35.11	<0.001
Urban	1735	253 (14.58)	1482 (85.42)
Rural	3301	711 (21.54)	2590 (78.46)
Smoking				0.76	0.385
No	2923	572 (19.57)	2351 (80.43)
Yes	2113	392 (18.55)	1721 (81.45)
Alcohol drinking				4.66	0.031
No	3189	640 (20.07)	2549 (79.93)
Yes	1847	324 (17.54)	1523 (82.46)
BMI (kg/m2)				51.89	<0.001
<18.5	244	89 (36.48)	155 (63.52)
18.5–23.9	2713	515 (18.98)	2198 (81.02)
≥24.0	2079	360 (17.32)	1719 (82.68)
Number of chronic diseases				24.85	<0.001
0	1758	275 (15.64)	1483 (84.36)
1	1670	330 (19.76)	1340 (80.24)
≥2	1608	359 (22.33)	1249 (77.67)

BMI, body mass index.

Longitudinal association of sleep duration with possible sarcopenia and its components

As shown in table 2, positive associations of shorter sleep duration (<6 hours) with possible sarcopenia (OR, 1.53; 95% CI, 1.31 to 1.78), low handgrip strength (OR, 1.69; 95% CI, 1.39 to 2.05) and low physical performance (OR, 1.43; 95% CI, 1.18 to 1.74) were observed in the unadjusted model when compared with the medium sleep duration (6–8 hours). In the fully adjusted model by gender, age, education, marital status, area of residence, smoking, alcohol drinking, BMI and number of chronic diseases, the associations remained significant for possible sarcopenia (OR, 1.22; 95% CI, 1.04 to 1.44) and low handgrip strength (OR, 1.27; 95% CI, 1.04 to 1.57) but not for low physical performance (OR, 1.19; 95% CI, 0.98 to 1.46). No statistical association was detected between longer sleep duration and the incidence of possible sarcopenia.

Table 2.

Longitudinal association of sleep duration with possible sarcopenia and its components

Outcomes	Sleep duration (OR, 95% CI)
	6–8 hours	<6 hours	>8 hours
Possible sarcopenia
Model 1	Reference	1.53 (1.31 to 1.78)	1.20 (0.92 to 1.56)
Model 2	Reference	1.27 (1.08 to 1.49)	1.04 (0.79 to 1.36)
Model 3	Reference	1.27 (1.08 to 1.49)	1.04 (0.79 to 1.36)
Model 4	Reference	1.22 (1.04 to 1.44)	1.03 (0.78 to 1.35)
Handgrip strength
Model 1	Reference	1.69 (1.39 to 2.05)	1.28 (0.92 to 1.78)
Model 2	Reference	1.38 (1.13 to 1.69)	1.08 (0.77 to 1.53)
Model 3	Reference	1.38 (1.13 to 1.69)	1.08 (0.77 to 1.53)
Model 4	Reference	1.27 (1.04 to 1.57)	1.03 (0.72 to 1.46)
Physical performance
Model 1	Reference	1.43 (1.18 to 1.74)	1.10 (0.79 to 1.53)
Model 2	Reference	1.20 (0.99 to 1.47)	0.96 (0.69 to 1.35)
Model 3	Reference	1.20 (0.99 to 1.46)	0.96 (0.69 to 1.35)
Model 4	Reference	1.19 (0.98 to 1.46)	0.97 (0.69 to 1.36)

CI, confidence interval; Model 1, no confounders were included; Model 2, gender, age, education, marital status and area of residence were included; Model 3, smoking and alcohol drinking were added into model 2; Model 4, BMI and number of chronic diseases were added into model 3; OR, odds ratio.

Exposure-response relation of sleep duration with possible arcopenia and its components

As shown in figure 2, a non-linear relationship between sleep duration and the incidence of possible sarcopenia was observed in the restricted cubic spline model. Shorter sleep duration was associated with a higher incidence of possible sarcopenia, while longer sleep duration was not.

Figure 2.

The exposure-response curve relations of sleep duration with possible sarcopenia and its components (models were adjusted by gender, age, education, marital status, area of residence, smoking, alcohol drinking, BMI and number of chronic diseases). BMI, body mass index.

Subgroup analysis

As shown in figure 3, a significant association of sleep duration with the incidence of possible sarcopenia in the male population was observed but not for the female population. In male, shorter and longer sleep durations were both associated with a higher incidence of possible sarcopenia, with ORs of 1.28 (95% CI: 1.00 to 1.62) and 1.48 (95% CI: 1.01 to 2.16). When stratified by age, the shorter sleep duration was associated with a higher incidence of possible sarcopenia in the 45–59-year group (OR, 1.30; 95% CI, 1.03 to 1.63).

Figure 3.

Forest plot for subgroup analysis by gender and age (except for stratified variable, models were adjusted by gender, age, education, marital status, area of residence, smoking, alcohol drinking, BMI and number of chronic diseases).

Discussion

This study analysed data from the 2011 baseline and 2015 follow-up survey of CHARLS. The longitudinal analysis based on nationally representative data among middle-aged and older adults in China suggested that individuals with shorter sleep duration were more likely to develop new-onset possible sarcopenia and low handgrip strength after 4 years of follow-up. In the fully adjusted regression model, participants who slept <6 hours per night had a 1.2-fold higher risk of developing possible sarcopenia compared with those who slept 6–8 hours per night, while participants who slept >8 hours per night had no significant risk of developing possible sarcopenia. Furthermore, the plot of restricted cubic splines exhibited an atypical inverse J-shaped association between sleep duration and possible sarcopenia.

This association may be due to adverse effects derived from short sleep duration on immune and endocrine systems. Previous studies have reported that sleep can regulate GH, cortisol, testosterone, insulin-like growth factor 1 (IGF-1), and insulin levels, which can re-establish muscle fibre, strength and function by mediating several protein synthesis and degradation pathways and favour the establishment of a highly proteolytic environment.^{23 24} A review examined the inflammatory/endocrine pathways that were affected by sleep loss, which may consequently interact with the GH/IGF-1 axis.²⁵ Meanwhile, insufficient sleep will accelerate the loss of muscle mass and hinder muscle recovery.²⁴ Previous findings also have indicated that sleep deprivation may result in increased secretion of proinflammatory cytokines.^{26 27} Inflammatory cytokines have been shown to prompt muscle wasting, ultimately stimulating protein catabolism and suppressing muscle synthesis.²⁸

In addition, sleep is accompanied by a marked increase in melatonin release. The circadian rhythm disruption and changes in melatonin with short sleep duration might be a potential underlying mechanism to explain the relationship between sleep duration and possible sarcopenia. Circadian rhythms play a crucial role in the structure and function of skeletal muscle as well as muscle metabolism, thereby being indispensable for the maintenance of skeletal muscle.²⁹ Melatonin plays an important role in regulating the circadian rhythm and sleep-wake cycle. And melatonin can protect mitochondria in skeletal muscle cells, maintain the number of muscle fibres, reverse the pathological changes in muscle tissue and improve the strength of skeletal muscle.³⁰ Higher levels of melatonin secretion were significantly associated with increased grip strength in older persons.³¹ Normalising circadian rhythms and sleep homeostasis may represent an appropriate strategy to preserve or recover muscle health.

Additionally, insufficient sleep may dysregulate metabolism through an altered gut microbiome. The host circadian rhythms influence microbial oscillations and composition, and in turn, microbial circadian oscillations impact host metabolism and immunity.³² Sleep duration was positively correlated with gut microbial diversity, and insufficient sleep could alter the total microbiome diversity, which was related to IL-6.³³ Higher levels of IL-6 were significantly associated with lower handgrip strength and muscle mass.³⁴ In older adults, short sleep duration was associated with an increase in proinflammatory bacteria.³⁵ In addition, gut microbiota may play a role in the development of sarcopenia.³⁶ Micronutrients and metabolites derived from the gut microbiota can reach and act on muscles through the gut-muscle axis.^{37 38}

The other potential mechanism might be explained by telomere length. Sleep loss can cause telomere shortening. Conversely, adequate sleep duration was positively associated with telomere length.³⁹ Studies have shown that longer telomere length was associated with a slower decline in grip strength, and a positive association existed between telomere length and muscular fitness in older adults.⁴⁰ An insufficient amount of sleep can affect a person’s cognitive performance or exercise capacity and increase the risk of injury during exercise.⁴¹ Daytime tiredness and reduced concentration due to insufficient sleep may contribute to a decrease in physical activity. Older adults who engaged in insufficient physical activity were more likely to have a higher risk of developing sarcopenia.¹⁷ These findings indicate that it is necessary to maintain appropriate sleep duration to prevent the occurrence of possible sarcopenia.

Using diet management to improve sleep is a convenient and inexpensive strategy. Some nutritional components or their metabolites in the diet have been proved to be beneficial for sleep wellness.⁴² Physical activity programmes can reduce cortisol levels and improve sleep quality and are an effective strategy to tackle stress and sleep complaints in adults.⁴³ According to the meridian theory, vitality acupunch exercise can promote qi and blood circulation and has a positive effect on the sleep quality in older adults with possible sarcopenia.⁴⁴ Music-based intervention appears to be a promising strategy to improve sleep quality.⁴⁵ In addition, taking a nap is a good measure to counteract the negative consequences of insufficient sleep duration.⁴⁶ Moreover, it is necessary to avoid heavy meals close to bedtime, caffeine, smoking and evening screen exposure.⁴⁷

Previous studies have reported gender differences in the relationship between sleep duration and the incidence of sarcopenia.^{11 15 48} In our study, we found a significant association between both shorter and longer sleep durations in men and an increased likelihood of possible sarcopenia; however, no such association was observed among women. This may be due to sex-related differences in muscle mass, sex hormones, employment status and social behaviour.⁴⁹ In addition, subgroup analysis indicated that middle-aged people with shorter sleep duration exhibited a higher likelihood of developing possible sarcopenia. Middle-aged people tend to have higher work pressure, poor sleep quality and later sleep timing.⁴⁹ Older adults with both shorter and longer sleep durations tended to develop possible sarcopenia, although the difference was not statistically significant. The excessive duration of sleep in older persons may be attributed to prolonged time spent in bed but with poor sleep quality, or it could indicate a decompensation of their physiological condition.¹²

Muscle strength and physical performance are two representative components of possible sarcopenia and health assessment indicators. Previous studies have disclosed that older adults with short sleep duration were associated with low muscle strength and had a significantly higher OR of having physical performance impairment.^{50 51} However, our study showed that short sleep duration was associated with low handgrip strength but not for physical performance. A recent study found that after adjusting for covariates, sleep duration was positively correlated with grip strength, but no significant association was observed with other physical function outcomes, which is consistent with our results.⁵²

There are several strengths of this study. First, the data we used come from a nationally representative longitudinal survey conducted in China. Thus, the results could be well representative. Second, this is the first study to examine the association between sleep duration and possible sarcopenia in China based on longitudinal analyses, and dose-response relationships are appraised. Third, this study provides new evidence for the utility of sleep duration in predicting the occurrence of possible sarcopenia, which is important for the prevention and early intervention of sarcopenia.

Nevertheless, there are a few limitations in this study. First, there is still the possibility of other unmeasured confounders that are not adequately considered in this study due to data limitations, such as physical activity, dietary intake and possible medications. Second, the assessment of sleep duration is measured using a self-reported questionnaire, so recall bias is inevitable. Furthermore, sleep duration is assessed merely with one single question; the use of sleep actigraphy may be more reliable. Third, CHARLS has amounts of missing or incomplete data which may lead to potential biases. Finally, although the results come from longitudinal analyses, causality remains unexplained. Further randomised controlled trials that objectively evaluate sleep-related indicators are needed to confirm the current findings. In addition, the current results cannot directly support the potential underlying mechanisms of the association between short sleep duration and possible sarcopenia.

Conclusion

In conclusion, this study suggested a correlation between short sleep duration and the occurrence of possible sarcopenia in the middle-aged Chinese population. Furthermore, there was a nonlinear dose-response relationship between sleep duration and possible sarcopenia. Enhancing sleep duration is beneficial for reducing and delaying the development of sarcopenia. And early interventions targeting obtaining an optimal sleep duration may help prevent a decline in muscle strength and reduce the burden of sarcopenia in China.

Data availability statement

Data are available in a public, open access repository. The CHARLS datasets generated and analysed during the current study are available in the website of the CHARLS home page at http://charls.pku.edu.cn/en. The CHARLS data are deidentified. Respondents are identified by a unique ID number.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants. The CHARLS survey project was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-11015). Participants gave informed consent to participate in the study before taking part.