Abstract
Background
Weakened handgrip strength is now a major public health concern. The relationship between body mass index (BMI) and handgrip strength (HGS) is controversial and inconclusive. We conducted cross-sectional and longitudinal analyses to investigate the association between BMI and HGS in middle-aged and older Chinese population.
Methods
We conducted a population-based cohort study using nationally representative data from the China Health and Retirement Longitudinal Study (CHARLS). A total of 5741 participants aged ≥ 60 years old at the baseline survey of CHARLS (2011) were included and a total of 2877 participants without low HGS were followed up in 2015. Multivariate linear and logistic regression models were constructed to evaluate the associations of BMI with HGS.
Results
The BMI was positively associated with HGS (β = 0.11; 95% CI: 0.07–0.15). During the 4 years of follow-up, 912 cases (31.7%) with low HGS were identified. Underweight BMI (OR: 1.56; 95% CI: 1.15–2.12) was significantly associated with increased risk of low HGS compared to the normal weight, while overweight (OR: 0.67; 95% CI: 0.55–0.83) and obesity (OR: 0.67; 95% CI: 0.50–0.91) were significantly associated with reduced risk of low HGS. Such associations were robust in subgroup analysis. Significant non-linear dose–response relationship between BMI and low HGS was also observed.
Conclusions
Our findings revealed that underweight BMI might be a predictor of low HGS in older adults, representing an important at-risk group for screening and intervention, and maintaining a higher BMI might prevent the development of low HGS in this population.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12877-025-06366-x.
Keywords: Handgrip strength, Body mass index, Ageing, Longitudinal study, CHARLS
Background
One of the primary challenges confronting healthcare organizations today is addressing the demands posed by an aging population [1]. Understanding the intricate relationship between body composition and physical function among elderly individuals is crucial for fostering healthy aging and mitigating age-related decline. The age-related decline in muscle strength is increasingly acknowledged as a significant contributor to the global burden of disease and disability, a burden that has been escalating in recent decades [2, 3]. Data from the Korean Longitudinal Study of Aging (KLoSA) revealed that the prevalence of low muscle strength was 13.0% in individuals under 65 years of age, escalating to 67.5% among those aged ≥ 65 years [4].
Hand grip strength (HGS), a pivotal indicator for evaluating muscle strength, has been demonstrated to be a convenient, cost-effective, and dependable measure in comprehensive physical assessments [5]. Numerous studies have demonstrated that grip strength is closely associated with the onset and progression of various chronic diseases and multiple health outcomes, including cardiovascular disease [6–8], liver diseases [9], diabetes mellitus [10], dementia [11], and mental health disorders [12]. A decline in HGS may indicate an elevated risk of developing these diseases, thus serving as a critical foundation for disease prediction and assessment [13]. Furthermore, monitoring alterations in HGS enables the prediction of patient mortality, quality of life, and cancer progression, thereby offering robust support for clinical decision-making [14–16].
Body Mass Index (BMI), widely accepted as a measure of adiposity and nutritional status, offers valuable insights into the prevalence of obesity, malnutrition, and associated chronic diseases among populations [17]. Previous studies have explored the relationship between BMI and HGS. Bim et al. observed a positive correlation between BMI and HGS in adolescents, with lean mass serving as a significant mediator in this relationship [18]. A British cross-sectional study conducted among middle-aged and elderly populations demonstrated that higher BMI is correlated with stronger HGS [19]. However, several studies have yielded conflicting results, suggesting that higher BMI is associated with lower HGS [20]. Thus far, the association between BMI and HGS remains contentious, warranting further investigation.
Thus, to gain deeper insights into the relationship between BMI and HGS, we conducted this study to investigate both the cross-sectional and longitudinal associations between BMI and the risk of low HGS among community-dwelling older adults aged 60 years and above, utilizing nationally representative data from the China Health and Retirement Longitudinal Study (CHARLS). This age group is particularly important to study as they are most vulnerable to sarcopenia, functional decline, and associated health complications. Additionally, age, gender, marital status, and education level were stratified to assess the magnitude of these associations.
Methods
Study population and data selection
China Health and Retirement Longitudinal Study (CHARLS) is an ongoing nationwide survey, providing an open-access database on socio-demographic, socioeconomic status, and health- related information with representative sample of the middle-aged and elderly population aged ≥ 45 years residing in local communities across China (150 counties/districts and 450 villages/resident committees from 28 provinces). The interviews in CHARLS were conducted in respondents’ homes by trained interviewers using a face-to-face computer-assisted personal interview (CAPI) method. The baseline survey was conducted in 2011 and 17,705 residents were included, with the follow-up surveys in 2013, 2015, 2018 and 2020, respectively. The program CHARLS was approved by the ethics committees of Peking University according to the 1964 Declaration of Helsinki. The informed consent was obtained from each participant before the investigation. A detailed description of the CHARLS design has been reported previously [21]. Our study followed all applicable specifications and guidelines of CHARLS.
CHARLS tracking data from baseline wave 2011 and wave 2015 were used in this study. The absence of BMI and HGS data in the 2018 and 2021 waves limits our ability to conduct a longer-term follow-up analysis, so the 2018 and 2021 data were not utilized. The independent variables and covariates were identified using data from baseline wave 2011, while the status of the dependent variable was identified using data from 2015. Of the 17,705 observations in wave 2011, a total of 5741 individuals were included in the cross-sectional analysis. Specific exclusion criteria were as follows: (1) age less than 60 years or missing data; (2) missing data for gender, education and residence; (3) lost to follow-up in wave 2015; (4) no available data for HGS in 2015. Finally, a total of 2877 subjects were included in the final analysis (Fig. 1).
Fig. 1.
Flow diagram for participants included in the study. Abbreviation: BMI, body mass index; HGS, handgrip strength
Assessment of HGS
HGS was measured in kilograms by trained volunteers using a YuejianTM WL-1000 dynamometer (Nantong Yuejian Physical Measurement Instrument Co., Ltd., Nantong, China) [21]. Subjects stood and started the test using either the dominant or non-dominant hand while receiving verbal encouragement. Each subject held the ergometer at a right angle (90°) and squeezed the handle for a few seconds, taking two measurements on the right and left hand. Participants were asked to provide maxi mum effort to perform the measurements [22]. The right and left-hand values were added and divided by 4 to calculate the average value. In accordance with the 2019 guidelines proposed by the Asian Working Group for Sarcopenia (AWGS), 28 kg and 18 kg were defined as cut-off points for low HGS for male and female, respectively [23].
Assessment of BMI
Height and weight information will be asked during the sur vey. Participants’ weights and heights were measured with shoes off and in light clothing, and the BMI is computed as follows: weight (kg) divided by height (m2). According to the overweight and obesity guidelines for Chinese adults [24], participants were grouped into four categories based upon their BMI: 1) underweight (< 18.5 kg/m2), 2) normal weight (18.5–23.9 kg/m2), 3) overweight (24–27.9 kg/m2), and 4) obesity (≥ 28 kg/m2).
Covariates
Demographic characteristics were collected using a questionnaire that included age, gender, income, marital status, education, and residence. Participants were categorized into 2 groups according to the age (group1: ≥ 60 years and < 70 years, group2: ≥ 70 years). Gender was divided into male and female. The marital status was divided into married and others. We classified the level of education into illiterate, primary, second/high school or above. The residence was grouped as an urban or rural area. Chronic diseases included: hypertension, diabetes, self-reported physician-diagnosed dyslipidemia and CVD. Subjects were diagnosed as hypertensive when the systolic blood pressure was ≥ 140 mmHg and the diastolic pressure was ≥ 90 mmHg or if antihypertensive medications were currently used. Diabetes was defined based on the use of insulin or oral hypoglycaemic agents, and plasma glucose ≥ 200 mg/dL.
Statistical analysis
Continuous variables were expressed as mean and standard deviation (SD). Categorical variables were expressed by Frequency and percentages, and chi-square tests were used to look for group differences.
By using multivariate linear and logistic regression analysis, the relationships between BMI and HGS were estimated in the cross-sectional and longitudinal analyses. BMI was treated as continuous variable or four-category variable and the normal weight group was used as a control. Three models were estimated: in model 1, unadjusted; in model 2, age, gender, marital status, residence, and education levels were adjusted; model 3 was adjusted as model 2 with further adjustment for smoking status, drinking status, hypertension, dyslipidemia, diabetes, and CVD. To better understand how these subgroup factors and BMI interact, we also performed subgroup analyses based on age, gender, marital status, residence, and education. Interaction analysis was performed to identify the effect modifications of sociodemographic characteristics, health related behaviors, and anthropometric measurements in the relationships between BMI and low HGS, by conducting likelihood ratio tests [25]. Moreover, the restricted cubic splines were used to examine possible non-linear associations and visualize the dose–response association of BMI with low HGS.
In all analyses, a two- tailed p < 0.05 was taken as a statistically significant difference. All statistical analyses were completed using R software (version 4.4.2). Restricted cubic splines were completed with the “rms” package.
Results
Baseline characteristics
The study participants were divided into four groups based on BMI, with 593 in the underweight group, 3130 in the normal weight group, 1477 in the overweight group, and 541 in the obesity group. There were 2884 men in total, accounting for 50.2% of the total sample of 5741 participants. The participants had a mean age of 68.1 years (SD 6.69), with 63.8% in the 60–69 age group and 36.2% aged 70 or above. Between the various BMI groups, significant differences were detected in terms of age, gender, marital status, residence, education, smoking status, drinking status, waist circumference (WC), glucose, triglycerides, HDL-C, LDL-C, hypertension, dyslipidemia, diabetes mellitus, cardiovascular disease (CVD), and handgrip strength (HGS) at baseline (all P values < 0.001) (Table 1).
Table 1.
Baseline characteristics of participants in this study according to BMI groups
| Characteristics | Overall | Underweight | Normal weight | Overweight | Obesity | p-Value |
|---|---|---|---|---|---|---|
| N = 5741 | N = 593 | N = 3130 | N = 1477 | N = 541 | ||
| Age(years) | 68.1 (6.69) | 71.1 (7.44) | 68.3 (6.74) | 67.0 (6.08) | 66.8 (5.96) | < 0.001 |
| Age group | < 0.001 | |||||
| 60–69 | 3665 (63.8%) | 271 (45.7%) | 1957 (62.5%) | 1050 (71.1%) | 387 (71.5%) | |
| 70 + | 2076 (36.2%) | 322 (54.3%) | 1173 (37.5%) | 427 (28.9%) | 154 (28.5%) | |
| Gender | < 0.001 | |||||
| Female | 2857 (49.8%) | 297 (50.1%) | 1380 (44.1%) | 834 (56.5%) | 346 (64.0%) | |
| Male | 2884 (50.2%) | 296 (49.9%) | 1750 (55.9%) | 643 (43.5%) | 195 (36.0%) | |
| Marital | < 0.001 | |||||
| Married | 4489 (78.2%) | 411 (69.3%) | 2413 (77.1%) | 1227 (83.1%) | 438 (81.0%) | |
| Others | 1252 (21.8%) | 182 (30.7%) | 717 (22.9%) | 250 (16.9%) | 103 (19.0%) | |
| Residence | < 0.001 | |||||
| Rural | 5265 (91.7%) | 577 (97.3%) | 2927 (93.5%) | 1301 (88.1%) | 460 (85.0%) | |
| Urban | 476 (8.29%) | 16 (2.70%) | 203 (6.49%) | 176 (11.9%) | 81 (15.0%) | |
| Education | < 0.001 | |||||
| Illiterate | 2140 (37.3%) | 285 (48.1%) | 1171 (37.4%) | 485 (32.8%) | 199 (36.8%) | |
| Primary | 2596 (45.2%) | 246 (41.5%) | 1469 (46.9%) | 659 (44.6%) | 222 (41.0%) | |
| High school or above | 1005 (17.5%) | 62 (10.5%) | 490 (15.7%) | 333 (22.5%) | 120 (22.2%) | |
| Smoking status | < 0.001 | |||||
| Non-smoker | 3269 (57.4%) | 305 (52.3%) | 1632 (52.5%) | 952 (64.8%) | 380 (70.4%) | |
| Ex-smoker | 665 (11.7%) | 62 (10.6%) | 345 (11.1%) | 189 (12.9%) | 69 (12.8%) | |
| Smoker | 1765 (31.0%) | 216 (37.0%) | 1130 (36.4%) | 328 (22.3%) | 91 (16.9%) | |
| Drinking status | < 0.001 | |||||
| Never | 3989 (69.5%) | 435 (73.4%) | 2069 (66.1%) | 1059 (71.8%) | 426 (78.7%) | |
| Former | 77 (1.34%) | 8 (1.35%) | 41 (1.31%) | 19 (1.29%) | 9 (1.66%) | |
| Current | 1673 (29.2%) | 150 (25.3%) | 1020 (32.6%) | 397 (26.9%) | 106 (19.6%) | |
| BMI (kg/m2) | 22.9 (4.00) | 17.1 (1.52) | 21.3 (1.52) | 25.7 (1.13) | 30.8 (3.91) | 0.000 |
| WC(cm) | 84.4 (12.9) | 71.6 (8.90) | 80.7 (9.78) | 91.4 (10.5) | 100 (13.0) | 0.000 |
| Glucose(mmol/l) | 112 (39.6) | 106 (28.8) | 110 (39.1) | 117 (43.9) | 119 (37.8) | < 0.001 |
| Triglycerides (mg/dL) | 128 (97.3) | 94.3 (46.7) | 116 (87.1) | 153 (116) | 164 (110) | < 0.001 |
| HDL-C (mg/dL) | 51.7 (15.7) | 60.4 (16.1) | 54.1 (15.9) | 46.3 (13.4) | 44.3 (12.2) | < 0.001 |
| LDL-C (mg/dL) | 118 (35.3) | 111 (33.9) | 115 (33.7) | 123 (37.0) | 124 (38.3) | < 0.001 |
| Hypertension | < 0.001 | |||||
| No | 3522 (61.3%) | 406 (68.5%) | 2026 (64.7%) | 827 (56.0%) | 263 (48.6%) | |
| Yes | 2219 (38.7%) | 187 (31.5%) | 1104 (35.3%) | 650 (44.0%) | 278 (51.4%) | |
| Dyslipidemia | < 0.001 | |||||
| No | 5095 (90.1%) | 563 (96.4%) | 2881 (93.5%) | 1246 (85.5%) | 405 (76.4%) | |
| Yes | 558 (9.87%) | 21 (3.60%) | 201 (6.52%) | 211 (14.5%) | 125 (23.6%) | |
| Diabetes mellitus | < 0.001 | |||||
| No | 5056 (88.2%) | 556 (93.8%) | 2830 (90.5%) | 1243 (84.4%) | 427 (78.9%) | |
| Yes | 679 (11.8%) | 37 (6.24%) | 298 (9.53%) | 230 (15.6%) | 114 (21.1%) | |
| CVD | < 0.001 | |||||
| No | 4851 (84.6%) | 528 (89.2%) | 2707 (86.6%) | 1213 (82.2%) | 403 (74.6%) | |
| Yes | 882 (15.4%) | 64 (10.8%) | 419 (13.4%) | 262 (17.8%) | 137 (25.4%) | |
| HGS (kg) | 26.1 (9.55) | 22.8 (8.55) | 26.2 (9.45) | 26.9 (9.87) | 26.7 (9.57) | < 0.001 |
| Low HGS (2011) | < 0.001 | |||||
| No | 3916 (68.2%) | 302 (50.9%) | 2086 (66.6%) | 1109 (75.1%) | 419 (77.4%) | |
| Yes | 1825 (31.8%) | 291 (49.1%) | 1044 (33.4%) | 368 (24.9%) | 122 (22.6%) |
BMI Body mass index, WC Waist circumference, HDL-C High density lipoprotein cholesterol, LDL-C Low-density lipoprotein cholesterol, CVD Cardiovascular disease, HGS Handgrip strength
Table S1 showed the characteristics of participants grouped by HGS status. Compared to normal HGS participants, individuals with low HGS were more likely to be older (71.0 vs 66.7 years), less educated (49.0% vs 31.8% illiterate), have other marital status (29.7% vs 18.1%), leaner (BMI 22.1 vs 23.3 kg/m2), underweight (15.9% vs 7.71%), live in rural settings (94.0% vs 90.6%), never drink (73.8% vs 67.5%), and more likely to have hypertension (42.2% vs 37.0%). Those with normal HGS were more likely to be overweight (28.3% vs 20.2%) or obese (10.7% vs 6.68%), have dyslipidemia (10.7% vs 8.00%), and have higher waist circumference, triglycerides, and LDL levels (all P values < 0.05).
Cross-sectional associations of BMI with HGS
In the cross-sectional study, BMI was positively associated with the continuous HGS in all models (Model 1: β = 0.09, P < 0.001; Model 2: β = 0.10, P < 0.001; Model 3: β = 0.11, P < 0.001) (Table S2). The prevalence of low HGS in total populations, underweight, normal weight, overweight and obesity individuals were 49.1% (291/593), 33.4% (1044/3130), 24.9% (368/1477), 22.6% (122/541), respectively (Table 1).
After adjustment for socio-demographic characteristics and health-related factors (Model 3), underweight (Odds Ratio, OR 1.54; 95% Confidence Interval, 95%CI, 1.27–1.87) was significantly associated with higher odds of low HGS (P < 0.001, P for trend < 0.001, Table 2), overweight (OR, 0.76; 95%CI, 0.65–0.88) and obesity (OR, 0.64; 95%CI, 0.51–0.81) were significantly associated with lower odds of low HGS (P < 0.001, P for trend < 0.001, Table 2).
Table 2.
ORs and 95% CIs of low HGS in the cross-sectional analysis
|
Model 1 OR(95%CI) |
p-Value |
Model 2 OR(95%CI) |
p-Value |
Model 3 OR(95%CI) |
p-Value | |
|---|---|---|---|---|---|---|
| BMI group | ||||||
| Normal weight | Ref | Ref | Ref | |||
| Under weight | 1.93 [1.61–2.30] | < 0.001 | 1.53 [1.27–1.85] | < 0.001 | 1.54 [1.27–1.87] | < 0.001 |
| Over weight | 0.66 [0.58–0.76] | < 0.001 | 0.77 [0.66–0.89] | < 0.001 | 0.76 [0.65–0.88] | < 0.001 |
| Obesity | 0.58 [0.47–0.72] | < 0.001 | 0.68 [0.54–0.85] | 0.015 | 0.64 [0.51–0.81] | < 0.001 |
| p for trend | < 0.001 | < 0.001 | < 0.001 | |||
Model 1: unadjusted
Model 2: adjusted for age, gender, marital status, residence, education
Model 3: adjusted for all the factors in Model 2 and smoking status, drinking status, hypertension, dyslipidemia, diabetes mellitus, and CVD
Longitudinal associations of baseline BMI with HGS at follow-up, 2011–2015
During the 4 years of follow-up, 912 cases (31.7%) with low HGS were identified. Table 3 shows the relationship between baseline BMI and the risk of low HGS.
Table 3.
Risk of low HGS according to baseline BMI groups, 2011 − 2015
|
Model 1 OR(95%CI) |
p-value |
Model 2 OR(95%CI) |
p-value |
Model 3 OR(95%CI) |
p-value | |
|---|---|---|---|---|---|---|
| BMI group | ||||||
| Normal weight | Ref | Ref | Ref | |||
| Under weight | 1.82 [1.36–2.43] | < 0.001 | 1.58 [1.17–2.14] | 0.003 | 1.56 [1.15–2.12] | 0.004 |
| Over weight | 0.67 [0.56–0.81] | < 0.001 | 0.71 [0.58–0.86] | < 0.001 | 0.67 [0.55–0.83] | < 0.001 |
| Obesity | 0.77 [0.58–1.01] | 0.063 | 0.77 [0.58–1.03] | 0.078 | 0.67 [0.50–0.91] | 0.010 |
| p for trend | < 0.001 | < 0.001 | < 0.001 | |||
Model 1: unadjusted
Model 2: adjusted for age, gender, marital status, residence, education
Model 3: adjusted for all the factors in Model 2 and smoking status, drinking status, hypertension, dyslipidemia, diabetes mellitus, and CVD
In the model 1, results showed that the underweight individuals had a higher risk of low HGS compared to normal weight group (OR, 1.82; 95% CI, 1.36–2.43). The results remained statistically significant in model 2 (OR, 1.58; 95% CI, 1.17–2.14) and model 3 (OR, 1.56; 95% CI, 1.15–2.12). However, the results in model 1 suggested that participants in overweight group had a decreased risk of low HGS compared to normal weight individuals (OR, 0.67; 95% CI, 0.56–0.81), while the obesity group showed a marginally significant decreased risk (OR, 0.77; 95% CI, 0.58–1.01, p = 0.063). The results for overweight remained statistically significant in model 2 (OR, 0.71; 95% CI, 0.58–0.86) and model 3 (OR, 0.67; 95% CI, 0.55–0.83), and the obesity group became statistically significant in model 3 (OR, 0.67; 95% CI, 0.50–0.91, p = 0.010). The P for trend was statistically significant across all models (all P for trend < 0.001, Table 3).
In Fig. 2, we used restricted cubic spline to model and visualize the dose–response relationship of BMI and the risk of low HGS. A non-linear relationship between BMI and risk of low HGS was observed (p for nonlinearity = 0.0015) (Fig. 2A). When stratified by sex, the non-linear association remained significant in males (p for nonlinearity = 0.0201) (Fig. 2B) but was marginally non-significant in females (p for nonlinearity = 0.0644) (Fig. 2C), though both showed overall significant associations (p < 0.001 and p = 0.0073, respectively). The curves for both sexes demonstrated a similar pattern, with a steeper increase in risk at lower BMI values. Age-stratified analyses revealed that the non-linear association was more pronounced in the 60–69 years old group (p for nonlinearity = 0.0009) (Fig. 2D) compared to those aged 70 + years (p for nonlinearity = 0.3752) (Fig. 2E), though both groups showed overall significant associations (p < 0.001 and p = 0.0299, respectively). Moreover, the longitudinal analysis showed a consistent positive association between BMI and continuous HGS (Model 1: β = 0.03, P = 0.07; Model 2: β = 0.0810, P < 0.001; Model 3: β = 0.10, P < 0.001) (Table S3).
Fig. 2.
Restricted cubic spline (RCS) of the associations between BMI and the risk of low HGS in the total included participants (A), in different sex groups (B and C), and in different age groups (D and E). The model was adjusted for age, gender, marital status, residence, education level, smoking status, drinking status, hypertension, dyslipidemia, diabetes mellitus, and CVD. BMI, body mass index; HGS, handgrip strength
Subgroup analysis of the associations between BMI and HGS
Subgroup analysis revealed consistent associations between BMI categories and low HGS across different demographic groups. In gender-stratified analyses, underweight males showed a significantly higher risk of low HGS compared to those with normal weight across all models (Model 3: OR, 2.09; 95% CI, 1.34 − 3.26), while this association was not statistically significant in females (Model 3: OR, 1.17; 95% CI, 0.76 − 1.79). Both overweight males (Model 3: OR, 0.71; 95% CI, 0.52 − 0.96) and females (Model 3: OR, 0.63; 95% CI, 0.58 − 0.83) had significantly lower risk of low HGS. Obesity was associated with significantly reduced risk of low HGS in males (Model 3: OR, 0.52; 95% CI, 0.30 − 0.89) but not in females (Model 3: OR, 0.75; 95% CI, 0.51 − 1.08) (Fig. 3). When examining subgroups by age, the underweight 60–69 years group showed significantly higher risk of low HGS (Model 3: OR, 1.71; 95% CI, 1.16 − 2.51), while both overweight (Model 3: OR, 0.71; 95% CI, 0.56 − 0.90) and obesity (Model 3: OR, 0.68; 95% CI, 0.48 − 0.97) were protective. In the 70 + years group, being underweight increased the risk of low HGS, though less dramatically (Model 3: OR, 1.36; 95% CI, 0.83 − 2.22), while overweight remained protective (Model 3: OR, 0.57; 95% CI, 0.39 − 0.85) (Fig. 4). The interaction test indicated that no significant difference was observed for the associations between gender or age with BMI categories and low HGS (p for interaction = 0.149 and 0.516, respectively), suggesting that the observed patterns were consistent across these demographic groups. Table S4 presented the results of subgroup analyses in rural and urban, in married and others, and in different education levels groups. Consistent positive associations between BMI and continuous HGS were observed in both male and female populations (Table S5 and S6).
Fig. 3.
ORs and 95% CIs for associations of BMI with low HGS risk stratified by gender. P inter: interaction tests between BMI and gender
Fig. 4.
ORs and 95% CIs for associations of BMI with low HGS risk stratified by age. P inter: interaction tests between BMI and age
Discussion
In this extensive population-based cohort study, we observed a protective effect of BMI, wherein higher BMI was correlated with a decreased risk of developing low HGS when BMI was treated as a continuous variable (not grouped). Upon stratifying by BMI, we observed that the underweight group exhibited a significantly elevated risk of developing low HGS compared to the normal weight group, while the overweight and obese groups displayed a reduced risk of developing low HGS. Through dose–response relationship analysis, we identified a non-linear correlation between BMI and the risk of low HGS. This study controlled for sociodemographic characteristics (age, gender, marital status, education, and residence), lifestyle-related indicators (drinking status, smoking status), and health-related variables (chronic diseases: hypertension, diabetes, dyslipidemia, CVD.
Our study contributes valuable insights to the existing evidence by investigating the longitudinal association between BMI and HGS in a cohort of older adults aged 60 years and above. Previous studies have documented a positive correlation between BMI and HGS across populations of all ages. For example, a positive relationship between BMI and HGS was observed in 118 adolescents (60 girls) aged 10–14 years, with lean mass serving as the mediator [18]. Moreover, robust evidence indicates that higher BMI from childhood onwards is correlated with stronger grip strength at age 46 years in both males and females, as demonstrated in a population-based study of over 7000 individuals followed from birth for almost five decades [26]. An analysis from the Brazilian Longitudinal Study of Ageing (ELSI-Brazil) revealed an inverse association between BMI and weak HGS [27]. The European Prospective Investigation into Cancer-Norfolk study demonstrated cross-sectional associations between higher BMI and stronger grip strength among adults aged 48–92 years [19]. Additionally, 8 UK cohort studies involving participants aged 50 to 90 + who contributed to the Healthy Ageing across the Life Course (HALCyon) research programme yielded consistent results in male participants, whereas no similar associations were observed in female participants [28]. Our findings align with the aforementioned studies, demonstrating positive associations in both male and female elderly populations. In our cross-sectional study, BMI was positively associated with HGS in both male and female populations. This relationship remains robust even after accounting for sociodemographic characteristics and health-related factors. Furthermore, we examined the association between different BMI subgroups and the risk of low HGS. The associations between BMI subgroups and the risk of low HGS remained consistent in both cross-sectional and longitudinal cohort studies. Our results suggest that the risk of low HGS is significantly elevated in the underweight group compared to the normal weight group, whereas it is slightly decreased in the overweight and obesity group. The dose–response relationship also indicated that as BMI increases, the risk of low HGS decreases in the underweight group. Moreover, our results showed age-stratified differences between the 60–69 and 70 + age groups, with the non-linear association between BMI and low HGS being more pronounced in the younger older adults (60–69 years) compared to those aged 70 + years. This suggests that interventions targeting BMI might be more effective in the early stages of aging.
Our findings are at odds with or diverge from several studies. In a cross-sectional analysis of 5783 community-dwelling adults from Wave 6 of the English Longitudinal Study of Ageing (ELSA), older adults with underweight BMI exhibited a 2.25-fold increased likelihood of probable sarcopenia compared to those within a healthy BMI range. However, their findings did not demonstrate a similar association between underweight BMI and low HGS (males: < 27 kg; females: < 16 kg); instead, significant relationships were observed between overweight and obese BMI and reduced risk of low HGS [29]. Another large population-based study that utilized longitudinal data and explored cumulative associations in Finnish adults aged over 55 years revealed that longer exposure to obesity throughout adulthood was associated with weaker grip strength [30]. A study conducted among 155 institutionalized and community-dwelling Chilean adults aged 65 years and older demonstrated that participants with obesity had a 3.2 times greater risk of presenting with sarcopenia compared to those with healthy nutritional status (defined by BMI) [31]. The inconsistency of associations can be influenced by various factors and necessitates careful consideration. Curtis et al. [29] reported a non-significant association between underweight BMI and low HGS (OR, 1.30; 95% CI, 0.65–2.62), although the association was positive. In the Finnish study [30], weight history was retrospectively self-reported, potentially explaining our contrasting findings. Crovetto Mattassi [31] et al. similarly observed a non-significant but positive relationship between underweight BMI and sarcopenia (OR, 7.82; 95% CI, 0.91–67.21) in the Chilean study, with the relatively small sample size potentially contributing to the lack of significance of the results. Additionally, the criteria for determining low HGS varied across studies. We also need to carefully consider the complexity of associations between BMI and HGS, which may be influenced by various factors such as the age and ethnicity of the participants, the period and location in which the study was conducted, among others.
One potential explanation for the associations between higher BMI and reduced risk of low HGS is that underweight may serve as a marker of malnutrition in older adults [32, 33], signaling weight loss and a decline in muscle mass. The findings of Akazawa et al.’s study, indicating that higher BMI is associated with greater muscle mass and less intramuscular adipose tissue in the quadriceps of hospitalized older patients, provided supportive evidence [34]. Muscle mass in men may be associated with androgen levels. Chu et al. emphasized that serum bioavailable testosterone levels decrease with aging and may contribute to central obesity and poorer muscle strength in aging men [35]. Additionally, fat mass increases as BMI rises, particularly in overweight or obese individuals. It acts as a mechanical load that elicits anabolic responses that promote muscle growth [36]. The anabolic response is typically greater in males than in females due to higher circulating levels of testosterone. This may account for the observed gender differences that female have a higher risk of developing low HGS than male. Several studies suggest that dynamic changes in BMI, such as weight loss or fluctuations, are closely associated with adverse outcomes in older adults, impacting cognitive function [37], physical performance [38], and mortality risk [39, 40]. Weight loss leads not only to loss of fat but also loss of muscle lean mass and bone [41]. Obviously, muscle lean mass loss is a major cause of sarcopenia which includes loss of walking speed and/or grip strength [42, 43]. Harris et al. showed that Greater variability in BMI is associated with declines in physical performance and higher rates of incident disability among older adults [38]. These findings underscore the need for future research to explore the longitudinal relationship between BMI dynamics and HGS in depth.
Because BMI does not differentiate between lean muscle and fat, it cannot accurately distinguish the fat content of individuals of different sexes. Cooper [26] et al.’s study observed higher BMI in males than females at any given age, with BMI increasing over time. This suggests that BMI is more likely to reflect higher levels of muscle mass and greater muscle mass accumulation. Undoubtedly, there is considerable value in employing more valid and accurate measures of body composition that distinguish between fat mass and lean muscle. The European Working Group on Sarcopenia in Older People in 2018(EWGSOP2) [44] and the Asian Working Group for Sarcopenia (AWGS) [23] recommend that dual-energy X-ray absorptiometry (DXA), computed tomography (CT), and magnetic resonance imaging (MRI) are more accurate measures of muscle mass. However, the widespread utilization of these methods is restricted for various reasons.
BMI is an easily measurable, readily available, inexpensive, and widely used indicator of health. Despite the limitations of BMI, we opted to investigate the relationship between BMI and HGS in older adults aged 60 and above to gather practical evidence for predicting or assessing the occurrence of low HGS.
The present study has several strengths. It is based on a large, prospective, robust, nationally representative sample of community-dwelling older adults in China. The large sample size and the use of standardized methods for measuring HGS enhanced the reliability of the results. We conducted both cross-sectional and longitudinal studies, considering BMI as both a continuous and categorical variable, to comprehensively investigate the relationship between BMI and HGS. The longitudinal association between BMI and HGS was well established, suggesting a causal relationship between BMI and HGS, as well as a predictive role of BMI in the occurrence of low HGS. The findings provided additional evidence to previous studies, indicating that underweight BMI significantly increased the risk of low HGS, while an increase in BMI could reduce this risk.
Limitations of this study are recognized. Although the study adjusted for relevant variables, there were still several factors that were not considered, such as diet and genetics. Only subjects with complete data in CHARLS were included in this study, potentially introducing selection bias. Furthermore, there is a good chance that participants’ BMI may have changed between 2011 and 2015. Changes in BMI may influence the association between BMI and HGS and should be taken into account in future studies. In addition, inclusion of data from the 2018 and 2021 waves would have more fully investigated the long-term relationship between BMI and HGS and strengthened the robustness of the conclusions. However, the absence of BMI and HGS data in the 2018 and 2021 waves limits our ability to conduct a longer-term follow-up analysis. This constraint restricts the exploration of potential long-term dynamic associations between BMI changes and HGS. We acknowledge this limitation and highlight the importance of future studies that incorporate additional follow-up data, should HGS measurements become available in subsequent waves. Such efforts would enable a more comprehensive investigation into the longitudinal relationships and provide deeper insights into critical periods for intervention.
Conclusions
In conclusion, our study of Chinese elderly adults aged 60 years and above yielded significant new evidence regarding the intricate associations between BMI and HGS, utilizing longitudinal data. The significant association between underweight and an increased likelihood of low HGS, coupled with the significant association between overweight and obesity and a reduced likelihood of low HGS, suggests that underweight BMI might serve as a predictor of low HGS, representing a crucial at-risk group for screening and intervention, while gaining BMI might mitigate the development of low HGS. Our findings suggest that maintaining adequate BMI is particularly important for preserving handgrip strength in older adults, with potentially stronger protective effects in the 60–69 age group compared to those aged 70 and above. Gender-specific differences in these associations further highlight the need for tailored approaches when considering BMI as a modifiable factor for preserving muscle strength in geriatric populations.
Supplementary Information
Acknowledgements
The data underlying the results of this study were obtained from the China Health and Retirement Longitudinal Study (CHARLS). The authors would like to express their gratitude to the CHARLS study team, the fieldwork team, and each respondent for their time and effort in contributing to the CHARLS project.
Authors’ contributions
HXX, SH and SMC conceived and designed the research; SH and SMC performed the data analysis; HXX, SH and HSX grafted the manuscript. WNZ and SMC critically revised the manuscript. All authors contributed to the interpretations of the findings. All authors reviewed the manuscript. The corresponding author had full access to all data in the study and assumed final responsibility for the decision to submit the manuscript for publication.
Funding
No external funding was provided for this study.
Data availability
The datasets analysed during the current study are available in the open CHARLS databases, https://charls.charlsdata.com/.
Declarations
Ethics approval and consent to participate
The CHARLS was approved by the Ethical Review Committee of Peking University (approval number: IRB00001052-11015), and carried out in accordance with the principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all participants.
Consent for publication
Not applicable.
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.
Haixia Xiao and Shan Huang contributed equally to this work.
References
- 1.Beard JR, Bloom DE. Towards a comprehensive public health response to population ageing. Lancet. 2015;385(9968):658–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sebbag E, Felten R, Sagez F, Sibilia J, Devilliers H, Arnaud L. The world-wide burden of musculoskeletal diseases: a systematic analysis of the World Health Organization burden of diseases database. Ann Rheum Dis. 2019;78(6):844–8. [DOI] [PubMed] [Google Scholar]
- 3.GBD Diseases and Injuries Collaborators. 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]
- 4.Jang SK, Kim JH, Lee Y. Effect of relative handgrip strength on cardiovascular disease among Korean adults aged 45 years and older: results from the Korean Longitudinal Study of Aging (2006–2016). Arch Gerontol Geriatr. 2020;86:103937. [DOI] [PubMed] [Google Scholar]
- 5.Labott BK, Bucht H, Morat M, Morat T, Donath L. Effects of exercise training on handgrip strength in older adults: a meta-analytical review. Gerontology. 2019;65(6):686–98. [DOI] [PubMed] [Google Scholar]
- 6.López-Bueno R, Andersen LL, Koyanagi A, Núñez-Cortés R, Calatayud J, Casaña J, Del Pozo Cruz B. Thresholds of handgrip strength for all-cause, cancer, and cardiovascular mortality: a systematic review with dose-response meta-analysis. Ageing Res Rev. 2022;82:101778. [DOI] [PubMed] [Google Scholar]
- 7.Liu G, Xue Y, Wang S, Zhang Y, Geng Q. Association between hand grip strength and stroke in China: a prospective cohort study. Aging. 2021;13(6):8204–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lu Y, Li G, Ferrari P, Freisling H, Qiao Y, Wu L, Shao L, Ke C. Associations of handgrip strength with morbidity and all-cause mortality of cardiometabolic multimorbidity. BMC Med. 2022;20(1):191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Petermann-Rocha F, Gray SR, Forrest E, Welsh P, Sattar N, Celis-Morales C, Ho FK, Pell JP. Associations of muscle mass and grip strength with severe NAFLD: a prospective study of 333,295 UK Biobank participants. J Hepatol. 2022;76(5):1021–9. [DOI] [PubMed] [Google Scholar]
- 10.Qiu S, Cai X, Yuan Y, Xie B, Sun Z, Wang D, Wu T. Muscle strength and prediabetes progression and regression in middle-aged and older adults: a prospective cohort study. J Cachexia Sarcopenia Muscle. 2022;13(2):909–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kuo K, Zhang YR, Chen SD, He XY, Huang SY, Wu BS, Deng YT, Yang L, Ou YN, Guo Y, et al. Associations of grip strength, walking pace, and the risk of incident dementia: a prospective cohort study of 340212 participants. Alzheimers Dement. 2023;19(4):1415–27. [DOI] [PubMed] [Google Scholar]
- 12.Jiang R, Westwater ML, Noble S, Rosenblatt M, Dai W, Qi S, Sui J, Calhoun VD, Scheinost D. Associations between grip strength, brain structure, and mental health in > 40,000 participants from the UK Biobank. BMC Med. 2022;20(1):286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yang Z, Wei J, Liu H, Zhang H, Liu R, Tang N, Yang X. Changes in muscle strength and risk of cardiovascular disease among middle-aged and older adults in China: evidence from a prospective cohort study. Chin Med J. 2024;137(11):1343–50. [DOI] [PMC free article] [PubMed]
- 14.Xie H, Ruan G, Wei L, Zhang H, Ge Y, Zhang Q, Lin S, Song M, Zhang X, Liu X, et al. Hand grip strength-based cachexia index as a predictor of cancer cachexia and prognosis in patients with cancer. J Cachexia Sarcopenia Muscle. 2023;14(1):382–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Dai KZ, Laber EB, Chen H, Mentz RJ, Malhotra C. Hand grip strength predicts mortality and quality of life in heart failure: insights from the Singapore cohort of patients with advanced heart failure. J Card Fail. 2023;29(6):911–8. [DOI] [PubMed] [Google Scholar]
- 16.Celis-Morales CA, Welsh P, Lyall DM, Steell L, Petermann F, Anderson J, Iliodromiti S, Sillars A, Graham N, Mackay DF, et al. Associations of grip strength with cardiovascular, respiratory, and cancer outcomes and all cause mortality: prospective cohort study of half a million UK Biobank participants. BMJ. 2018;361:k1651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Meshkat S, Liu Y, Jung H, Tassone VK, Pang H, Janssen-Aguilar R, Lou W, Bhat V. Temporal associations of BMI and glucose parameters with depressive symptoms among US adults. Psychiatry Res. 2024;332:115709. [DOI] [PubMed] [Google Scholar]
- 18.Bim MA, de Pinto AA, Claumann GS, Pelegrini A. Mediation effects of lean mass and fat mass on the relationship between body mass index and handgrip strength. Am J Hum Biol. 2023;36(4):e24004. [DOI] [PubMed]
- 19.Keevil VL, Luben R, Dalzell N, Hayat S, Sayer AA, Wareham NJ, Khaw KT. Cross-sectional associations between different measures of obesity and muscle strength in men and women in a British cohort study. J Nutr Health Aging. 2015;19(1):3–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Soraya N, Parwanto E. The controversial relationship between body mass index and handgrip strength in the elderly: an overview. Malays J Med Sci. 2023;30(3):73–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. 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]
- 22.Zhang L, Guo L, Wu H, Gong X, Lv J, Yang Y. Role of physical performance measures for identifying functional disability among Chinese older adults: data from the China health and retirement longitudinal study. PLoS ONE. 2019;14(4):e0215693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chen LK, Woo J, Assantachai P, Auyeung TW, Chou MY, Iijima K, Jang HC, Kang L, Kim M, Kim S, et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc. 2020;21(3):300-307.e302. [DOI] [PubMed] [Google Scholar]
- 24.Chen C, Lu FC. The guidelines for prevention and control of overweight and obesity in Chinese adults. Biomed Environ Sci. 2004;17 Suppl:1–36. [PubMed] [Google Scholar]
- 25.Wang YX, Shan Z, Arvizu M, Pan A, Manson JE, Missmer SA, Sun Q, Chavarro JE. Associations of menstrual cycle characteristics across the reproductive life span and lifestyle factors with risk of type 2 diabetes. JAMA Netw Open. 2020;3(12):e2027928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cooper R, Tomlinson D, Hamer M, Pinto Pereira SM. Lifetime body mass index and grip strength at age 46 years: the 1970 British cohort study. J Cachexia Sarcopenia Muscle. 2022;13(4):1995–2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.de Souza MB, de Souza Andrade AC, Lustosa Torres J, de Souza BL, de Carvalho BA, de Melo Mambrini JV, Lima-Costa MF. Nationwide handgrip strength values and factors associated with muscle weakness in older adults: findings from the Brazilian Longitudinal Study of Aging (ELSI-Brazil). BMC Geriatr. 2022;22(1):1005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hardy R, Cooper R, Aihie Sayer A, Ben-Shlomo Y, Cooper C, Deary IJ, Demakakos P, Gallacher J, Martin RM, McNeill G, et al. Body mass index, muscle strength and physical performance in older adults from eight cohort studies: the HALCyon programme. PLoS ONE. 2013;8(2):e56483. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Curtis M, Swan L, Fox R, Warters A, O’Sullivan M. Associations between body mass index and probable sarcopenia in community-dwelling older adults. Nutrients. 2023. 10.3390/nu15061505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Stenholm S, Sallinen J, Koster A, Rantanen T, Sainio P, Heliövaara M, Koskinen S. Association between obesity history and hand grip strength in older adults–exploring the roles of inflammation and insulin resistance as mediating factors. J Gerontol A Biol Sci Med Sci. 2011;66(3):341–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Crovetto Mattassi M, Henríquez Mella C, Pérez Bocaz L. Association between sarcopenia and nutritional status in Chilean older people aged 65 years and older. Nutrients. 2022. 10.3390/nu14245228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Cederholm T, Jensen GL, Correia M, Gonzalez MC, Fukushima R, Higashiguchi T, Baptista G, Barazzoni R, Blaauw R, Coats AJS, et al. Glim criteria for the diagnosis of malnutrition - a consensus report from the global clinical nutrition community. J Cachexia Sarcopenia Muscle. 2019;10(1):207–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Bardon LA, Streicher M, Corish CA, Clarke M, Power LC, Kenny RA, O’Connor DM, Laird E, O’Connor EM, Visser M, et al. Predictors of incident malnutrition in older Irish adults from the Irish Longitudinal Study on Ageing cohort-a MaNuEL study. J Gerontol A Biol Sci Med Sci. 2020;75(2):249–56. [DOI] [PubMed] [Google Scholar]
- 34.Akazawa N, Kishi M, Hino T, Tsuji R, Tamura K, Hioka A, Moriyama H. Higher body mass index in hospitalized older patients is related to higher muscle quality. J Nutr Health Aging. 2022;26(5):495–500. [DOI] [PubMed] [Google Scholar]
- 35.Chu LW, Tam S, Kung AW, Lo S, Fan S, Wong RL, Morley JE, Lam KS. Serum total and bioavailable testosterone levels, central obesity, and muscle strength changes with aging in healthy Chinese men. J Am Geriatr Soc. 2008;56(7):1286–91. [DOI] [PubMed] [Google Scholar]
- 36.Tomlinson DJ, Erskine RM, Morse CI, Winwood K, Onambélé-Pearson G. The impact of obesity on skeletal muscle strength and structure through adolescence to old age. Biogerontology. 2016;17(3):467–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wang RT, Sun Z, Tan CC, Tan L, Xu W. Dynamic features of body mass index in late life predict cognitive trajectories and Alzheimer’s disease: a longitudinal study. Journal of Alzheimers Dis. 2024;100(4):1365–78. [DOI] [PubMed] [Google Scholar]
- 38.McMenamin KJ, Harris TB, Baker JF. Weight variability, physical functioning and incident disability in older adults. J Cachexia Sarcopenia Muscle. 2023;14(4):1648–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Chen MJ, Lai YJ, Chen CC, Hsieh C, Chou YC, Yen YF. Modification of risk for all-cause and cardiovascular disease-related mortality with changes in the body mass index in older individuals: a population-based cohort study. Gerontology. 2024;70(9):903–13. [DOI] [PubMed] [Google Scholar]
- 40.Sun Q, Xia X, He F. Longitudinal association between body mass index (BMI), BMI trajectories and the risk of frailty among older adults: a systematic review and meta-analysis of prospective cohort studies. Arch Gerontol Geriatr. 2024;124:105467. [DOI] [PubMed] [Google Scholar]
- 41.Huo YR, Suriyaarachchi P, Gomez F, Curcio CL, Boersma D, Gunawardene P, Demontiero O, Duque G. Comprehensive nutritional status in sarco-osteoporotic older fallers. J Nutr Health Aging. 2015;19(4):474–80. [DOI] [PubMed] [Google Scholar]
- 42.Masanés F, Rojano ILX, Salvà A, Serra-Rexach JA, Artaza I, Formiga F, Cuesta F, López Soto A, Ruiz D, Cruz-Jentoft AJ. Cut-off points for muscle mass - not grip strength or gait speed - determine variations in sarcopenia prevalence. J Nutr Health Aging. 2017;21(7):825–9. [DOI] [PubMed] [Google Scholar]
- 43.Balogun S, Winzenberg T, Wills K, Scott D, Jones G, Aitken D, Callisaya ML. Prospective associations of low muscle mass and function with 10-year falls risk, incident fracture and mortality in community-dwelling older adults. J Nutr Health Aging. 2017;21(7):843–8. [DOI] [PubMed] [Google Scholar]
- 44.Cruz-Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyère O, Cederholm T, Cooper C, Landi F, Rolland Y, Sayer AA, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48(1):16–31. [DOI] [PMC free article] [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 datasets analysed during the current study are available in the open CHARLS databases, https://charls.charlsdata.com/.