Waist‐to‐height ratio and renal dysfunction as independent correlates of low handgrip strength in patients with type 2 diabetes

ABSTRACT

Introduction

Handgrip strength (HGS) is a key indicator of functional decline. We investigated whether central adiposity (waist‐to‐height ratio, WHtR) and renal dysfunction independently or additively correlate with low HGS across body mass index (BMI) categories in Asian patients with type 2 diabetes.

Materials and Methods

This retrospective cross‐sectional study included 1,468 patients. Low HGS was defined by the Asian Working Group for Sarcopenia 2019 criteria. Multivariable logistic regression evaluated independent and combined correlates. Subgroup analyses examined the interactions between WHtR, BMI, and chronic kidney disease (CKD).

Results

Older age, higher albuminuria, CKD (adjusted odds ratio [aOR] = 1.98), and WHtR >0.5 (aOR = 1.82) were independently associated with low HGS, whereas BMI > 24 kg/m² was protective (aOR = 0.44). Significant effect modification existed between BMI and WHtR (P for interaction = 0.026): central obesity neutralized the protective effect of higher BMI (aOR shifted from 0.13 to 0.81). Furthermore, while WHtR >0.5 and CKD showed no multiplicative interaction, their co‐occurrence exerted a clear additive clinical burden, yielding the highest risk for low HGS (aOR = 1.74, P = 0.025) compared to patients without either condition.

Conclusions

Central obesity and renal dysfunction (impaired eGFR and albuminuria) are distinct, additive correlates of low HGS. WHtR effectively unmasks functional vulnerability, especially in normal‐weight individuals. Integrating WHtR and comprehensive renal parameters provides a robust, low‐cost screening strategy for sarcopenia risk in diabetes care.

In 1,468 patients with type 2 diabetes, central obesity (WHtR >0.5) neutralizes the protective effects of higher BMI on muscle strength. Concurrently, visceral adiposity and CKD impose an additive threat, maximizing low handgrip strength risk (aOR = 1.74). Combining WHtR and renal tracking optimizes clinical sarcopenia screening.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder that affects multiple organ systems and predisposes individuals to various complications beyond hyperglycemia. Among these complications, loss of muscle strength and function, commonly referred to as sarcopenia, has emerged as a severe but underdiagnosed comorbidity ¹ , ² . Sarcopenia in type 2 diabetes mellitus results from a complex interplay of insulin resistance, low‐grade chronic inflammation, neuropathy, nutritional deficiencies, aging, and iatrogenic factors ² , ³ , ⁴ , ⁵ . This condition is associated with adverse outcomes, such as falls, frailty, functional disability, poor quality of life, and increased mortality.

Handgrip strength (HGS), a simple, noninvasive, and widely validated measure of upper‐limb muscle strength, serves as a practical screening tool for sarcopenia and physical frailty in both clinical and epidemiological settings ⁶ , ⁷ . A decline in HGS is not only an early indicator of sarcopenia but also a marker of broader systemic deterioration ⁶ , ⁷ , ⁸ , ⁹ . However, the extent to which central adiposity and renal dysfunction independently or jointly affect muscle strength remains insufficiently characterized in Asian populations.

Abdominal obesity, characterized by visceral fat accumulation, contributes to systemic inflammation and insulin resistance, both of which are key drivers of muscle catabolism ¹⁰ , ¹¹ . Compared with body mass index (BMI), waist‐to‐height ratio (WHtR) is a sensitive and age‐ and sex‐independent marker of abdominal obesity and cardiometabolic risk ¹² . Chronic kidney disease (CKD) drives muscle loss through mechanisms such as oxidative stress, uremic toxin accumulation, insulin resistance, and impaired protein turnover ¹³ . However, the combined effects of WHtR and CKD markers on muscle strength in type 2 diabetes mellitus have not been comprehensively investigated. The present study addressed the research gap by conducting a retrospective analysis of data from a large clinical cohort of patients with type 2 diabetes mellitus in Taiwan. We specifically evaluated the independent association of WHtR and renal function with low HGS and analyzed the interaction between WHtR and BMI.

METHODS

Study design and participants

This retrospective cross‐sectional study included 1,468 patients with type 2 diabetes mellitus who attended the metabolism clinic at Kaohsiung Chang Gung Memorial Hospital. Patients were eligible for inclusion in this study if their HGS data were available. The exclusion criteria were as follows: neuromuscular diseases, severe cognitive impairment, inability to comply with standardized testing protocols, and incomplete clinical data. The study protocol was approved by the Institutional Review Board of Kaohsiung Chang Gung Memorial Hospital (approval number: 202400219B0). The requirement for informed consent was waived because of the retrospective study design.

Data collection

Data were extracted from electronic medical records and outpatient clinic databases. Demographic and anthropometric variables included age, sex, type 2 diabetes mellitus duration, height, weight, BMI, waist circumference, and WHtR.

BMI was analyzed both as a continuous variable and categorical variable by using a cut‐off of 24 kg/m² to define overweight, following Taiwan Ministry of Health and Welfare standards ¹⁴ . This cut‐off aligns with the World Health Organization recommendations for Asian populations, acknowledging that Asian populations accumulate fat and develop metabolic risks at lower BMI levels than do Western populations ¹⁵ . A WHtR of >0.5 is indicative of excessive central obesity and visceral fat accumulation. It predicts cardiometabolic risk more accurately than does BMI alone ¹² .

Laboratory variables included glycated hemoglobin (HbA1c), total cholesterol, triglycerides, low‐density lipoprotein cholesterol (LDL‐C), and high‐density lipoprotein cholesterol (HDL‐C). Renal function was evaluated using serum creatinine, estimated glomerular filtration rate (eGFR), and urinary albumin‐to‐creatinine ratio (UACR).

We further analyzed behavioral factors, medication use, and type 2 diabetes mellitus‐related complications—for example, current or former smoking and alcohol use, exercise habits (aerobic and resistance), use of major antidiabetic agents (metformin, sulfonylureas, dipeptidyl peptidase‐4 inhibitors, sodium‐glucose cotransporter‐2 inhibitors, acarbose, pioglitazone, glucagon‐like peptide‐1 receptor agonists, or insulin), use of lipid‐lowering and antihypertensive medications, and presence of cardiovascular disease (atherosclerotic cardiovascular disease, peripheral arterial occlusive disease, acute coronary syndrome, or heart failure). Diabetic retinopathy was assessed through fundoscopic examination, and neuropathy was assessed through nerve conduction studies or 10‐g monofilament testing. CKD was defined as an eGFR of <60 mL/min/1.73 m².

Measurement of muscle strength

Handgrip strength (kilograms) was measured using a TTM digital hand dynamometer (TTM‐YD; Tokyo, Japan), a Smedley‐type device widely used in Asian epidemiological studies ¹⁶ . Measurements were performed with participants standing, with the arms naturally hanging at the sides and the elbows fully extended, following standardized protocols for Smedley dynamometers ¹⁶ , ¹⁷ . The dominant hand was tested twice, with a rest interval of at least 1 min between trials, and the highest value was recorded for analysis.

Measurement of waist circumference

Waist circumference was measured using a flexible tape at the midpoint between the iliac crest and the lowest rib. The participant remained standing with their arms relaxed on both sides and the trunk free of clothing. Measurements were obtained with the abdomen relaxed at the end of expiration ¹⁸ . In the present study, WHtR was used as a practical and validated anthropometric surrogate for visceral adiposity.

Variable definitions

Low HGS was defined on the basis of the 2019 Asian Working Group for Sarcopenia (AWGS) consensus as a HGS of <28 kg for men and < 18 kg for women ⁷ .

Statistical analysis

Continuous variables are presented as mean ± standard deviation or median (interquartile range) values, depending on data distribution. Categorical variables are presented as counts and percentages. Data normality was assessed. Skewed variables, including UACR, were log‐transformed before analysis. Intergroup comparisons were performed using Student's t‐test or the Mann–Whitney U test for continuous variables, and the chi‐square test for categorical variables, as appropriate.

Univariate logistic regression analyses were conducted to identify factors associated with low HGS. Variables demonstrating clinical relevance or statistical significance in univariate analyses were entered into multivariate logistic regression models by using the Enter method. Multicollinearity was assessed to avoid inclusion of highly correlated variables. Correlation coefficients were calculated using Pearson or Spearman methods, as appropriate. Variance inflation factors and tolerance were additionally examined. No problematic multicollinearity was observed (all variance inflation factors <5 and tolerances >0.2).

In subgroup analysis and interaction testing, we examined effect modification by central obesity. Participants were cross‐classified into four BMI × WHtR subgroups using predefined cutoffs (BMI > 24 vs ≤24 kg/m²; WHtR ≤0.5 vs >0.5). Individuals with a BMI of ≤24 kg/m² and a WHtR of ≤0.5 constituted the reference group. Crude odds ratios (OR) were calculated from event counts. Adjusted odds ratio (aOR) and 95% confidence interval (CI) were estimated using multivariable logistic regression with the 4‐level subgroup variable as a categorical predictor. Models were adjusted for age, diabetes duration, total cholesterol, LDL‐C, serum creatinine, ln‐transformed UACR, antihypertensive agent use, cardiovascular disease, retinopathy, and CKD. We then fitted a second model including BMI category, WHtR category, and a multiplicative interaction term (BMI × WHtR). Similarly, for WHtR × eGFR subgroup analysis, participants were stratified into four groups based on WHtR (cut‐off: 0.5) and eGFR (cut‐off: 60 mL/min/1.73 m²), with the WHtR ≤0.5 and eGFR ≥60 group as the reference. Adjusted ORs were calculated using a multivariable model controlling for the same covariates except for CKD. The Wald test was applied to evaluate the multiplicative interaction term. Statistical significance was set at P < 0.05. All analyses were performed using SPSS (version 27; IBM, Armonk, NY, USA).

RESULTS

A total of 1,468 participants with type 2 diabetes mellitus were included in the analysis and stratified by HGS status. Table 1 presents the baseline characteristics of these participants. The mean age of the overall cohort was 61.6 ± 11.8 years, and 55.1% of the participants were men. Of the total population, 959 (65.3%) participants had normal HGS, whereas 509 (34.7%) had reduced HGS. The reduced HGS group was significantly older than the normal HGS group (66.4 ± 10.7 vs. 59.1 ± 11.6 years, P < 0.001). Furthermore, type 2 diabetes mellitus duration was significantly longer in the reduced HGS group than in the normal HGS group (12.8 ± 8.0 vs. 10.7 ± 7.2 years, P < 0.001). Height, body weight, and BMI were all lower in the reduced HGS group than in the normal HGS group (all P < 0.001). The reduced HGS group was less likely than the normal HGS group to have a BMI of >24 kg/m² (64.4% vs. 75.6%, P < 0.001). Waist circumference was slightly lower in the reduced HGS group than in the normal HGS group (89.5 ± 10.0 vs. 91.3 ± 11.1 cm, P = 0.001). The proportion of participants with a WHtR of >0.5 was modestly higher in the reduced HGS group than in the normal HGS group (87.2% vs. 83.3%, P = 0.048). Sex distribution did not differ significantly between the two groups.

Table 1.

Baseline characteristics of study participants stratified by handgrip strength status

Variable	Overall (N = 1,468)	Normal HGS (n = 959)	Low HGS † (n = 509)	P‐value*
Demographics and anthropometrics
Age (years)	61.6 ± 11.8	59.1 ± 11.6	66.4 ± 10.7	<0.001***
Male	809 (55.1)	533 (55.6)	276 (54.2)	0.619
Diabetes duration (years)	11.4 ± 7.6	10.7 ± 7.2	12.8 ± 8.0	<0.001***
Height (cm)	161.5 ± 8.5	162.9 ± 8.6	158.9 ± 7.6	<0.001***
Weight (kg)	69.5 ± 14.1	71.7 ± 14.6	65.3 ± 12.2	<0.001***
BMI (kg/m2)	26.5 ± 4.4	26.9 ± 4.5	25.8 ± 4.2	<0.001***
BMI > 24	1,053 (71.7)	725 (75.6)	328 (64.4)	<0.001***
Waist circumference (cm)	90.7 ± 10.8	91.3 ± 11.1	89.5 ± 10.0	0.001***
WHtR >0.5	1,243 (84.7)	799 (83.3)	444 (87.2)	0.048*
Laboratory parameters
HbA1c (%)	7.29 ± 1.10	7.30 ± 1.11	7.28 ± 1.09	0.699
HbA1c ≥ 9	105 (7.2)	76 (7.9)	29 (5.7)	0.115
Total cholesterol (mg/dL)	164.8 ± 33.2	167.2 ± 33.7	160.3 ± 31.7	<0.001***
TG (mg/dL)	111 (81)	114 (86)	108 (73)	0.085
LDL‐C (mg/dL)	90.5 ± 27.3	92.4 ± 27.9	87.0 ± 25.8	<0.001***
HDL‐C (mg/dL)	48.6 ± 16.3	49.1 ± 17.9	47.7 ± 12.8	0.119
Creatinine (mg/dL)	1.00 ± 0.46	0.96 ± 0.41	1.06 ± 0.52	<0.001***
eGFR (mL/min/1.73 m2)	77.8 ± 25.4	80.2 ± 23.8	73.3 ± 27.6	<0.001***
UACR (mg/g)	14.8 (48.8)	12.9 (37.7)	21.5 (76.8)	<0.001***
Ln‐transformed UACR	3.16 ± 1.63	3.00 ± 1.56	3.45 ± 1.71	<0.001***
Lifestyle, medications, and comorbidities
Current or former smoker	112 (7.6)	81 (8.4)	31 (6.1)	0.106
Current or former drinker	56 (3.8)	43 (4.5)	13 (2.6)	0.066
Regular exercise	709 (48.3)	454 (47.3)	255 (50.1)	0.314
Aerobic exercise	697 (47.5)	444 (46.3)	253 (49.7)	0.213
Resistance exercise	16 (1.1)	12 (1.3)	4 (0.8)	0.414
Metformin	1,280 (87.2)	842 (87.8)	438 (86.1)	0.340
Sulfonylurea	739 (50.3)	469 (48.9)	270 (53.0)	0.131
DPP‐4i	694 (47.3)	436 (45.5)	258 (50.7)	0.056
SGLT2i	275 (18.7)	181 (18.9)	94 (18.5)	0.849
Acarbose	120 (8.2)	84 (8.8)	36 (7.1)	0.262
Pioglitazone	250 (17.0)	171 (17.8)	79 (15.5)	0.262
GLP‐1 RA	24 (1.6)	17 (1.8)	7 (1.4)	0.568
Premix insulin	115 (7.8)	69 (7.2)	46 (9.0)	0.211
Basal insulin	93 (6.3)	64 (6.7)	29 (5.7)	0.465
Rapid insulin	34 (2.3)	23 (2.4)	11 (2.2)	0.774
Lipid‐lowering agent	1,098 (74.8)	714 (74.5)	384 (75.4)	0.678
Antihypertensive agent	791 (53.9)	488 (50.9)	303 (59.5)	0.002**
Cardiovascular disease	203 (13.8)	115 (12.0)	88 (17.3)	0.005**
Retinopathy	637 (43.4)	378 (39.4)	259 (50.9)	<0.001***
Neuropathy	182 (12.4)	117 (12.2)	65 (12.8)	0.752
CKD ‡	307 (20.9)	151 (15.7)	156 (30.6)	<0.001***

Data are presented as mean ± standard deviation (SD), number (percentage), or median (interquartile range [IQR]) for skewed variables. BMI, body mass index; CKD, chronic kidney disease; DPP‐4i, dipeptidyl peptidase‐4 inhibitor; eGFR, estimated glomerular filtration rate; GLP‐1 RA, glucagon‐like peptide‐1 receptor agonist; HbA1c, glycated hemoglobin; HDL‐C, high‐density lipoprotein cholesterol; HGS, handgrip strength; LDL‐C, low‐density lipoprotein cholesterol; Ln‐UACR, Ln‐transformed urine albumin‐to‐creatinine ratio; SGLT2i, sodium‐glucose cotransporter 2 inhibitor; TG, triglycerides; UACR, urine albumin‐to‐creatinine ratio; WHtR, waist‐to‐height ratio. All tests were two‐sided.

^* P < 0.05.

^** P < 0.01.

^*** P < 0.001.

^† Low HGS was defined using the AWGS 2019 cutoffs (men <28 kg; women <18 kg).

^‡ CKD was defined as an eGFR of <60 mL/min/1.73 m².

Regarding laboratory findings, no significant between‐group differences were observed in HbA1c level or the proportion of participants with an HbA1c of ≥9%. The following parameters were significantly lower in the reduced HGS group than in the normal HGS group: total cholesterol (160.3 ± 31.7 vs. 167.2 ± 33.7 mg/dL, P < 0.001), LDL‐C (87.0 ± 25.8 vs. 92.4 ± 27.9 mg/dL, P < 0.001), and eGFR (73.3 ± 27.6 vs. 80.2 ± 23.8 mL/min/1.73 m², P < 0.001). By contrast, the following parameters were significantly higher in the reduced HGS group than in the normal HGS group: serum creatinine (1.06 ± 0.52 vs. 0.96 ± 0.41 mg/dL, P < 0.001) and UACR (median: 21.5 vs. 12.9 mg/g, P < 0.001).

Lifestyle factors such as smoking, alcohol use, and regular exercise were similar between the two groups and were not significantly associated with HGS status. No significant between‐group differences were observed in the use of antidiabetic medication. However, antihypertensive medication use was more frequent in the reduced HGS group than in the normal HGS group (59.5% vs. 50.9%, P = 0.002). The following type 2 diabetes mellitus‐related complications were significantly more prevalent in the reduced HGS group than in the normal HGS group: cardiovascular disease (17.3% vs. 12.0%, P = 0.005), diabetic retinopathy (50.9% vs. 39.4%, P < 0.001), and CKD (30.6% vs. 15.7%, P < 0.001). However, no significant between‐group difference was noted in the prevalence of neuropathy.

Univariate logistic regression analyses identified several factors significantly associated with reduced HGS. Older age (OR: 1.067) and longer type 2 diabetes mellitus duration (OR: 1.038) were associated with a higher risk of low HGS. Furthermore, the following renal markers were strongly associated with a higher risk of reduced HGS: higher serum creatinine level (OR: 1.591), lower eGFR (OR: 0.989), and higher UACR (ln‐transformed UACR; OR: 1.182). Among type 2 diabetes mellitus‐related conditions, antihypertensive medication use (OR: 1.420), cardiovascular disease (OR: 1.534), diabetic retinopathy (OR: 1.592), and CKD (OR: 2.365) were significantly associated with reduced HGS. By contrast, lower height, weight, BMI, and total cholesterol level as well as smaller waist circumference were inversely associated with reduced HGS. Additional univariate results are presented in Table 2.

Table 2.

Univariate and multivariate logistic regression analyses of factors associated with low HGS ^†

	Univariate		Multivariate
	OR	P‐value	aOR § (95% CI)	P‐value
Age	1.067	<0.001	1.056 (1.037–1.075)	<0.001***
Diabetes duration	1.038	<0.001	1.015 (0.992–1.038)	0.200
BMI > 24	0.582	<0.001	0.439 (0.292–0.661)	<0.001***
WHtR >0.5	1.368	0.048	1.820 (1.065–3.108)	0.028*
Total cholesterol	0.993	<0.001	0.988 (0.977–1.000)	0.052
LDL‐C	0.993	<0.001	1.009 (0.995–1.024)	0.212
Creatinine	1.591	<0.001	0.661 (0.394–1.108)	0.116
ln‐UACR	1.182	<0.001	1.118 (1.005–1.244)	0.040*
Antihypertensive agent	1.420	0.002**	0.924 (0.653–1.307)	0.655
Cardiovascular disease	1.534	0.005**	1.113 (0.824–1.602)	0.637
Retinopathy	1.592	<0.001	0.926 (0.658–1.303)	0.660
CKD ‡	2.365	<0.001	1.975 (1.167–3.343)	0.011*

All tests were two‐sided. BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; LDL‐C, low‐density lipoprotein cholesterol; ln, natural logarithm; OR, odds ratio; UACR, urinary albumin‐to‐creatinine ratio; WHtR, waist‐to‐height ratio.

^* P < 0.05.

^** P < 0.01.

^*** P < 0.001.

^† Low HGS was defined using the AWGS 2019 cutoffs (men <28 kg; women <18 kg).

^‡ CKD was defined as an eGFR of <60 mL/min/1.73 m².

^§ Model adjusted for age, diabetes duration, total cholesterol, LDL‐C, serum creatinine, Ln‐transformed UACR, antihypertensive agent use, cardiovascular disease, retinopathy, and CKD.

In multivariate logistic regression models adjusted for potential confounders, factors independently associated with low HGS included older age (aOR: 1.056; 95% confidence interval [CI], 1.037–1.075; P < 0.001), WHtR >0.5 (aOR: 1.820; 95% CI: 1.065–3.108; P = 0.028), higher ln‐transformed UACR (aOR: 1.118; 95% CI: 1.005–1.244; P = 0.040), and CKD (aOR: 1.975; 95% CI: 1.167–3.343; P = 0.011). By contrast, a BMI of >24 kg/m² was independently associated with a lower risk of reduced HGS (aOR: 0.439; 95% CI: 0.292–0.661; P < 0.001; Table 2).

Correlation analysis (Table 3) revealed that HGS was significantly and negatively correlated with age (r = −0.335, P < 0.001) and type 2 diabetes mellitus duration (r = −0.149, P < 0.001) and positively correlated with height (r = 0.669), body weight (r = 0.473), BMI (r = 0.138), and waist circumference (r = 0.273) (all P < 0.001). Regarding cardiometabolic parameters, HGS was negatively correlated with high‐density lipoprotein cholesterol level (r = −0.117) and UACR (r = −0.135; both P < 0.001). Regarding renal markers, HGS was positively associated with serum creatinine level (r = 0.101, P < 0.001) and eGFR (r = 0.061, P = 0.019). No significant correlations were observed between HGS and HbA1c percentage, total cholesterol level, triglyceride level, or LDL‐C level.

Table 3.

Correlation coefficients among continuous variables related to handgrip strength

Variable	Correlation coefficient	P‐value
Age	−0.335	<0.001
Diabetes duration	−0.149	<0.001
Height	0.669	<0.001
Weight	0.473	<0.001
BMI	0.138	<0.001
Waist circumference	0.273	<0.001
HbA1C	0.004	0.887
Total cholesterol	−0.011	0.662
TG	0.058	0.025
LDL‐C	0.043	0.101
HDL‐C	−0.117	<0.001
Creatinine	0.101	<0.001
eGFR	0.061	0.019
ln‐transformed UACR	−0.135	<0.001

Values represent Pearson or Spearman correlation coefficients, as appropriate. Abbreviations: BMI, body mass index; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; ln, natural logarithm; TG, triglycerides; UACR, urinary albumin‐to‐creatinine ratio.

Participants were further stratified into four groups by BMI (>24 vs. ≤24 kg/m²) and WHtR (>0.5 vs. ≤0.5). Individuals with a BMI of ≤24 kg/m² and a WHtR of ≤0.5 constituted the reference group. Participants with normal BMI but central obesity (BMI ≤24 kg/m² and WHtR >0.5) had a potentially higher risk of reduced HGS than did the reference group (aOR: 1.42, P = 0.106). By contrast, participants with a BMI of >24 kg/m² and a WHtR of ≤0.5 had a significantly lower risk of reduced HGS than did the reference group (aOR: 0.13, P = 0.001). Compared to the reference group, no significant difference in the risk of low HGS was observed in participants with both elevated BMI and central obesity (BMI > 24 kg/m² and WHtR >0.5; aOR: 0.81, P = 0.263). To formally evaluate effect modification, we included a multiplicative interaction term (BMI category × WHtR category) in the multivariable model; the interaction was statistically significant (P for interaction = 0.026) (Figure 1).

Figure 1.

Association between subgroups defined by body mass index (BMI) and waist‐to‐height ratio (WHtR) and the risk of low handgrip strength (HGS). Data are presented as n/N (%) and adjusted odds ratios (ORs) with 95% confidence intervals (CIs). Low HGS was defined based on the Asian Working Group for Sarcopenia (AWGS) 2019 criteria (men <28 kg; women <18 kg). ORs were adjusted for age, diabetes duration, total cholesterol, LDL‐C, serum creatinine, ln‐UACR, use of antihypertensive agents, history of cardiovascular disease, retinopathy, and chronic kidney disease (CKD). The interaction between BMI category and WHtR category was statistically significant (P for interaction = 0.026) in a multivariable logistic regression model including the multiplicative interaction term. Group 1 (BMI ≤24 kg/m² and WHtR ≤0.5) served as the reference group. All tests were two‐sided. *P < 0.05.

Furthermore, additional analyses were performed to evaluate the combined and interactive effects of central adiposity and renal dysfunction. There was no significant multiplicative interaction between WHtR >0.5 and CKD regarding the risk of low HGS (p for interaction = 0.277). However, when patients were stratified into four subgroups based on the presence of these conditions, those with concurrent central obesity (WHtR >0.5) and CKD exhibited a significantly higher risk of low HGS (adjusted OR = 1.741, P = 0.025) compared to the reference group with neither condition in the fully adjusted model (Figure 2).

Figure 2.

Association between subgroups defined by waist‐to‐height ratio (WHtR) and estimated glomerular filtration rate (eGFR) and the risk of low handgrip strength (HGS). Data are presented as n/N (%) and adjusted odds ratios (ORs) with 95% confidence intervals (CIs). Low HGS was defined based on the Asian Working Group for Sarcopenia (AWGS) 2019 criteria (men <28 kg; women <18 kg). ORs were adjusted for age, diabetes duration, total cholesterol, LDL‐C, serum creatinine, ln‐UACR, use of antihypertensive agents, history of cardiovascular disease, and retinopathy. The interaction between WHtR category and eGFR category was statistically insignificant (P for interaction = 0.277) in a multivariable logistic regression model including the multiplicative interaction term. Group 1 (WHtR ≤0.5 and eGFR ≧60) served as the reference group. All tests were two‐sided. *P < 0.05.

DISCUSSION

Superiority of WHtR: Unmasking central adiposity

This study identified the multifactorial correlates of low HGS in patients with type 2 diabetes mellitus. By examining a comprehensive set of demographic, anthropometric, and metabolic variables, we demonstrated that abdominal obesity (assessed using WHtR) and renal dysfunction (manifested as CKD and albuminuria) were independently associated with reduced muscle strength. These findings highlight the limitations of BMI alone and suggest that individuals with normal BMI but central obesity may represent a clinically vulnerable subgroup. A pivotal finding of the present study is the identification of WHtR >0.5 as a potent and independent predictor of low HGS. Although BMI is widely used to assess general obesity, it does not distinguish lean mass from adipose tissue and may therefore obscure cardiometabolic risk related to visceral fat ¹⁹ , ²⁰ , ²¹ , ²² . We adopted the cut‐off of 0.5 for WHtR in accordance with the 2025 Lancet Diabetes & Endocrinology Commission consensus, which recommends this threshold as a universal, age‐ and sex‐independent indicator of excess adiposity ²³ .

Wasit‐to‐height ratio is fundamentally a metric of central fat distribution rather than a direct index of muscle function. Therefore, deriving a cohort‐specific WHtR cut‐off to ‘diagnose’ low muscle strength would be inappropriate. Instead, our study utilizes the established 0.5 threshold to identify the presence of a lipotoxic and proinflammatory visceral fat environment. Our results suggest that once this threshold of adverse fat distribution is crossed, metabolic crosstalk significantly accelerates the deterioration of muscle function. Thus, WHtR serves as a marker of risk exposure rather than a direct diagnostic tool. This perspective aligns with the literature implicating visceral fat in the development of metabolic disturbances ¹⁹ , ²⁰ , ²¹ , ²² . Mechanistically, visceral adiposity is associated with increased production of proinflammatory cytokines, such as tumor necrosis factor‐α and interleukin‐6, which contribute to a state of chronic low‐grade inflammation that accelerates muscle catabolism ¹¹ , ²⁴ . Furthermore, ectopic lipid accumulation exacerbates insulin resistance, which impairs glucose uptake and protein synthesis in myocytes, ultimately contributing to sarcopenia ² , ²⁵ , ²⁶ , ²⁷ .

The obesity paradox: The protective role of lean mass

Our multivariate analysis revealed a negative association between low HGS and BMI > 24 kg/m², indicating a protective effect of increased body weight. This finding is consistent with the ‘obesity paradox’ frequently reported in patients with chronic diseases ²⁸ , ²⁹ . In the context of muscle strength, a higher BMI is often associated with greater absolute lean body mass, which is required to support increased body weight during daily activities ⁴ , ³⁰ , ³¹ . This sustained mechanical loading provides a continuous stimulus for muscle use, contributing to the preservation of muscle function and attenuation of age‐related decline ³² . However, our findings indicate that this protective benefit is conditional and heavily dependent on fat distribution, as indicated by the subgroup analysis.

Interplay of BMI and WHtR: The ‘counterbalancing’ effect

The stratification of participants by BMI and WHtR provided novel insights into the interaction between general and central obesity.

The ‘masked’ high risk (normal BMI, high WHtR)

In crude analyses, patients with a BMI ≤24 kg/m² but a WHtR >0.5 exhibited the highest odds of low HGS (OR = 1.90). This association was attenuated and became non‐significant after multivariable adjustment (aOR = 1.42, 95% CI 0.93–2.19; P = 0.106). Nevertheless, the point estimate remained elevated, suggesting a potentially vulnerable subgroup that may be overlooked when BMI is used alone. This finding is particularly relevant to Asian patients, who are predisposed to the ‘metabolically obese, normal‐weight’ phenotype ³³ , ³⁴ . Research suggests that, at a given BMI, Asian individuals have a higher proportion of total body fat and greater visceral adiposity than do Western individuals ³⁴ . Consequently, reliance on conventional BMI cut‐off alone may substantially underestimate both metabolic and sarcopenic risks in Asian populations. Our findings indicate that a considerable proportion of individuals classified as having normal body weight (BMI ≤24 kg/m²) had central obesity (WHtR >0.5), a phenotype that showed a consistently higher point estimate for low HGS. These results underscore the clinical value of WHtR as a complementary screening metric to identify the otherwise concealed vulnerability of Asian populations.

The ‘true’ low risk (high BMI, low WHtR)

Patients with a BMI of >24 but a WHtR of ≤0.5 had the lowest risk of reduced HGS (aOR: 0.13; P = 0.001). This pattern is consistent with the ‘metabolically healthy obesity’ phenotype, in which individuals exhibit preserved insulin sensitivity and muscle quality despite excess body weight ³⁵ . In such individuals, adipose tissue is predominantly stored in subcutaneous depots rather than visceral compartments. This subcutaneous fat acts as a safe ‘metabolic sink’ for fatty acids, protecting skeletal muscle from lipotoxicity and ectopic fat deposition ³⁶ , ³⁷ . In the absence of viscerally driven inflammatory stress, the mechanical loading associated with higher body weight predominates, contributing to the preservation of muscle strength.

The ‘counterbalancing’ effect (high BMI, high WHtR)

This protective association is markedly attenuated in the presence of central obesity. Compared to the reference group, participants with both elevated BMI and central obesity (BMI > 24 kg/m² and WHtR >0.5) showed no significant difference in the risk of low HGS (aOR: 0.81, P = 0.263). A comparison between individuals with a BMI of >24 kg/m² and a WHtR of >0.5 and those with a BMI of >24 kg/m² and a WHtR of ≤0.5 yielded a key insight: central obesity effectively attenuated the strong protective association observed in the metabolically healthier overweight group, with the aOR increasing from 0.13 to 0.81. This pattern reflects the interaction of two opposing physiological processes. Although higher body weight confers a mechanical stimulus that supports muscle mass (the ‘obesity paradox’) ³¹ , excessive visceral adiposity promotes a metabolic toxicity through inflammation and insulin resistance ¹¹ . The deleterious metabolic effects of visceral fat therefore neutralize the mechanical advantage of higher weight ³⁸ . These findings indicate that central obesity can shift a high‐BMI phenotype from one that is functionally robust to one that is functionally vulnerable, reinforcing the concept that higher body weight is advantageous for muscle strength only when visceral adiposity remains limited ³⁸ , ³⁹ , ⁴⁰ .

The kidney‐muscle axis: A dual threat

Our findings further highlighted the renal–muscle axis as a key correlate of diabetic sarcopenia. CKD (stage ≥3) and albuminuria emerged as strong, independent predictors of low HGS, indicating that renal dysfunction contributes to muscle deterioration through active pathophysiological processes rather than serving merely as a comorbid condition ⁴¹ , ⁴² . Possible mechanisms include uremic toxin accumulation, metabolic acidosis, oxidative stress, protein‐energy wasting, chronic systemic inflammation, and impaired clearance of advanced glycation end products, all of which are exacerbated in CKD ¹³ , ⁴¹ , ⁴³ . Notably, the independent association between albuminuria and reduced HGS highlights the role of systemic endothelial dysfunction and microvascular damage, which may compromise capillary supply to muscle fibers, reducing nutrient delivery and accelerating atrophy ⁴⁴ , ⁴⁵ , ⁴⁶ . The concurrent identification of impaired filtration capacity (CKD stage) and glomerular barrier dysfunction (albuminuria) as risk factors supports a multidisciplinary approach to managing sarcopenia risk.

Our subgroup analysis further elucidated the interplay between central adiposity and renal function. While no significant multiplicative interaction was observed, the co‐existence of WHtR >0.5 and CKD resulted in the highest prevalence (52.4%) and odds (aOR = 1.741) of low handgrip strength. This additive effect suggests that central obesity can further compromise functional integrity in patients who already have renal impairment, reinforcing the necessity of monitoring both WHtR and kidney health markers in diabetes care.

The cholesterol paradox and potential systemic confounders

Interestingly, our multivariable analysis revealed an inverse trend between total cholesterol levels and the risk of low handgrip strength (aOR = 0.988, P = 0.052). This finding highlights the clinical ‘cholesterol paradox’. A plausible and critical mediator for this association is thyroid dysfunction, particularly unrecognized hyperthyroidism. Excess circulating thyroid hormone accelerates lipid clearance, resulting in low serum total cholesterol, while simultaneously inducing skeletal muscle catabolism and thyrotoxic myopathy, which directly impairs grip strength ⁴⁷ . Furthermore, exceptionally low total cholesterol in patients with prolonged type 2 diabetes mellitus may also reflect an underlying state of malnutrition, chronic inflammation, or advanced frailty, which concurrently impedes muscle preservation ⁴⁸ , ⁴⁹ . Since thyroid function tests were not routinely evaluated in our dataset, future prospective studies incorporating comprehensive thyroid profiles are warranted to confirm these mechanistic pathways.

Comparisons with the literature and novelty

This study has several distinct contributions to the literature. Although sarcopenia has been extensively examined in older populations, very few studies have specifically analyzed the combined roles of WHtR and renal dysfunction in a large Asian cohort with type 2 diabetes mellitus. Unlike studies that rely solely on waist circumference, which necessitates sex‐specific thresholds, the use of WHtR offers a height‐adjusted and clinically practical index that is particularly suitable for Asian body proportions ⁵⁰ , ⁵¹ . Moreover, by explicitly investigating the interaction between BMI and WHtR, our analysis provides direct empirical support for the heightened vulnerability of the metabolically obese, normal‐weight phenotype with respect to muscle strength, a relationship that may be obscured in studies treating BMI and waist‐based measures as independent covariates rather than interacting variables ⁵² .

Strengths and limitations

The strengths of this study include its large sample size, which provided sufficient statistical power to perform detailed subgroup analyses of the interaction between BMI and WHtR. In addition, standardized HGS assessment with a digital dynamometer and comprehensive analysis of metabolic and renal variables enhanced the robustness and internal validity of our findings.

This study has several limitations. First, the cross‐sectional design precluded causal inference. Therefore, it is difficult to definitively determine whether renal dysfunction and central obesity accelerate HGS decline, or whether individuals with low HGS and frailty are more prone to developing CKD and adiposity. Prospective longitudinal studies are required to clarify directionality.

Second, although HGS is a validated and widely accepted proxy for overall muscle quality, direct assessments of muscle mass (e.g., dual‐energy X‐ray absorptiometry or bioelectrical impedance analysis) and physical performance (e.g., walking speed) were not performed. This limited our ability to formally diagnose sarcopenia according to the full AWGS criteria. Additionally, specific inflammatory markers (such as hs‐CRP or IL‐6) were unavailable in our real‐world retrospective dataset, meaning the proposed inflammatory mechanisms linking central obesity to muscle weakness remain speculative. Nevertheless, we emphasize that HGS is increasingly recognized as a powerful, standalone prognostic biomarker. The AWGS 2019 consensus explicitly validates the use of low HGS to identify ‘possible sarcopenia’ in primary care settings to facilitate early intervention ⁷ . Furthermore, global studies have demonstrated that HGS is a highly sensitive predictor of cardiometabolic risk and mortality, often outperforming muscle mass metrics because it captures overall neuromuscular functional quality ⁶ , ⁷ , ⁹ . In busy clinical settings where routine DXA scans and inflammatory profiling are impractical, HGS and WHtR serve as highly accessible, zero‐cost screening tools for the early triage of high‐risk patients.

Finally, because this was a single‐center retrospective analysis, residual confounding from unmeasured factors cannot be fully excluded. For instance, dietary protein intake and physical activity intensity were not routinely recorded. Despite these limitations, the large sample size and comprehensive metabolic evaluations bolster the internal validity of our findings.

CONCLUSION

Abdominal obesity and renal dysfunction are independent and complementary correlates of muscle weakness in type 2 diabetes mellitus. A WHtR of >0.5 may serve as a clinically informative screening tool to unmask sarcopenia risk, particularly in individuals with normal BMI who may otherwise be overlooked. The counterbalancing effect observed in individuals with overweight plus central obesity further challenges the assumption that higher body weight uniformly confers functional benefits. Routine assessment of WHtR and renal markers (CKD stage and albuminuria) should be incorporated into type 2 diabetes mellitus care to improve risk stratification and inform targeted interventions, such as resistance training and renal‐protective strategies, and thus preserve muscle function in this high‐risk population.

DISCLOSURE

The authors declare no conflicts of interest.

Approval of the research protocol: N/A.

Informed consent: N/A.

Registry and the registration no. of the study/trial: N/A.

Animal studies: N/A.

AUTHOR CONTRIBUTIONS

Study concept and design: Shing‐Hua Chen, Chen‐Kai Chou, Shao‐Wen Weng, Jung‐Fu Chen, Feng‐Chih Shen. Acquisition of data: Shing‐Hua Chen, Feng‐Chih Shen. Analysis and interpretation of data: Shing‐Hua Chen, Chen‐Kai Chou, Shao‐Wen Weng, Jung‐Fu Chen, Feng‐Chih Shen. Drafting of the manuscript: Shing‐Hua Chen, Chen‐Kai Chou, Feng‐Chih Shen. Critical revision of the manuscript for important intellectual content: Shao‐Wen Weng, Jung‐Fu Chen.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.