The association between grip strength and kidney stones: A National Health and Nutrition Examination Survey and Mendelian randomization study

Abstract

Objective

The formation of kidney stones is influenced by multiple factors, including physical health indicators such as grip strength (GS), which reflects overall muscle health. This study aimed to investigate the relationship between GS and the risk of developing kidney stones.

Methods

Data from the National Health and Nutrition Examination Survey between 2011 and 2014 were analyzed. The maximum value of six GS measurements was selected as the definitive value. Relative grip strength (rGS) was calculated to adjust for body size differences. Chi-square tests and t-tests were employed for categorical and continuous variables, respectively. Logistic regression models were used to evaluate the relationships of GS and rGS with the prevalence of kidney stones. Two-sample Mendelian randomization using the inverse-variance weighted method verified the causal association.

Results

The study included 9306 participants, with 302 having kidney stones. Participants with kidney stones had lower mean GS (32.0 kg vs. 37.3 kg, p<0.001) and rGS (1.12 vs. 1.34, p<0.001). Logistic regression showed that higher GS and rGS (divided into quartiles Q1–Q4) were associated with a reduced risk of kidney stones (GS: Q4 vs. Q1, odds ratio 0.32, 95% confidence interval 0.20–0.51, p<0.001; rGS: Q4 vs. Q1, odds ratio 0.45, 95% confidence interval 0.28–0.71, p<0.001). Two-sample Mendelian randomization results confirmed a negative association between right-hand GS and the risk of kidney stones.

Conclusion

Lower GS is linked to a higher risk of kidney stones, highlighting the importance of muscle health in the prevention of kidney stones.

1. Introduction

Kidney stone disease, a common urological condition, has been noted to exhibit an increasing prevalence globally, posing significant challenges to public health and healthcare systems [1]. The etiology of kidney stones is multifaceted, involving a complex interplay of genetic predispositions, dietary habits, obesity, metabolic syndrome, and physical inactivity, among other clinical factors (such as underlying health conditions) [2]. The gender and ethnic differences are also notable in prior research [3]. Especially, the increase in obesity, diabetes, salt intake and animal protein intake, high consumption of sweetened beverages, global warming, and sedentary behavior may lead to an increase in the incidence and prevalence of kidney stones [[4], [5], [6]]. The associated lifestyle and physical examination factors highlight the crucial role that regular physical activity, including muscle-strengthening exercises, and healthy lifestyle choices play in the prevention and management of kidney stones.

Sarcopenia, marked by a gradual decline in skeletal muscle mass and strength [7], has emerged as an additional factor that may influence the risk of developing kidney stones [8,9]. In a National Health and Nutrition Examination Survey (NHANES) study, sarcopenia was confirmed to be positively associated with kidney stone prevalence through a sarcopenia index [8]. A randomized clinical trial also reflected the association between muscular strength and kidney function, which underlies the potential lithogenic factors [10].

Grip strength (GS), a simple and non-invasive measure of overall muscular health, serves as a critical indicator for assessing sarcopenia [11]. However, GS has not been utilized as a variable to study its relationship with the risk of kidney stones to date. Using GS as an indicator of sarcopenia and exploring its relationship with nephrolithiasis may be a worthwhile topic for further research, although the nature and directionality of this relationship remain to be fully elucidated.

In this study, we aimed to leverage the NHANES database to investigate the relationship between GS and kidney stone prevalence. Furthermore, we sought to employ Mendelian randomization (MR) techniques to infer causality. The MR results are less susceptible to residual confounding and reverse causality bias, as genetic variants are randomly distributed during meiosis and are generally uncorrelated with environmental factors [12]. By combining these methods, this study aimed to investigate the potential associations between GS and kidney stone formation.

2. Materials and methods

2.1. NHANES sample participants

The data utilized in this analysis were sourced from the NHANES datasets conducted between 2011 and 2014. NHANES is conducted biennially to assess the health and nutritional status of the non-institutionalized United States population. NHANES is renowned for its large sample size and systematic multistage sampling design, which allow the database to be representative of the civilian and non-institutionalized United States population. Given these advantages, we selected this database. The study protocol received approval from the Institutional Review Board of the National Center for Health Statistics (Ethics Review Board protocol number: Continuation of Protocol #2011–17), and informed consent was obtained from all participants in the original NHANES study. The methodology and data from NHANES are publicly available online at https://www.cdc.gov/nchs/nhanes/index.htm.

The researchers included 19 931 participants from NHANES in this study. Participants were excluded based on the following criteria: (a) unknown kidney stone status (n=8616); (b) unknown GS levels (n=1420); (c) unknown education levels (n=5); (d) unknown marital status (n=3); (e) unknown hypertension status (n=10); (f) unknown diabetes mellitus status (n=5); (g) unknown body mass index (BMI) (n=0); (h) unknown renal function status (n=562); (i) unknown physical activity status (n=4).

2.2. NHANES GS assessment

The GS measurements for participants were obtained from the “Examination Data” section of the NHANES dataset under the “Muscle Strength–Grip Test” (https://wwwn.cdc.gov/nchs/data/nhanes/public/2013/manuals/Muscle_Strength_2013.pdf). During the assessment, each participant's GS was recorded three times for each hand. For the purposes of statistical analysis, the maximum value out of these six trials was selected as the definitive measurement.

A prior study indicated that relative grip strength (rGS), calculated by dividing GS by BMI, showed a stronger association with cardiovascular and metabolic disease biomarkers compared with maximum GS [13]. Conversely, when it came to predicting tumor outcomes, the maximum GS demonstrated superior performance over rGS [14]. Consequently, both rGS and absolute GS were utilized to evaluate the muscular strength of the participants.

2.3. NHANES study variables

All information regarding study variables was extracted from the NHANES database, encompassing details such as age, gender, ethnicity, marital status, education level, BMI, hypertension status, diabetes mellitus status, smoking status, physical activity level, GS, blood urea nitrogen, creatinine, and uric acid levels. The clinical attributes of the participants were organized as follows: age (<40 years and ≥40 years), race (non-Hispanic White, non-Hispanic Black, Mexican American, other Hispanic, and other), marital status (married and unmarried), education level (less than high school, high school or equivalent, and college or above), BMI (underweight, normal, overweight, and obese), hypertension (yes or no), diabetes mellitus (yes, no, or borderline), smoking status (never, former, or current smoker), physical activity status (vigorous [yes or no] and moderate [yes or no]), and ever had kidney stones (yes or no).

2.4. NHANES data analysis

To compare those with kidney stones with those without, a chi-square test was used to calculate the p-values for categorical variables, while a t-test for coefficients was conducted within generalized linear models for continuous outcomes.

Participants were categorized into four groups based on the quartiles of the rGS distribution (Q1–Q4) [15]. This classification allowed for an examination of how the included variables might influence rGS across different quartile groups. Because NHANES has a national multistage probability sampling design, weights were included in the analysis. “WTMEC2YR” (weight for the mobile examination center examined sample for 2 years) was selected as the primary weighting variable, and other weighting variables included “SDMVPSU” (sampling units) and “SDMVSTRA” (stratified sampling weights).

To examine the association between the levels of GS and rGS and the risk of kidney stones, logistic regression models were utilized. Odds ratios (ORs) and 95% confidence intervals (CIs) for the prevalence of kidney stones associated with the quartiles of GS and rGS were calculated. Five logistic regression models were conducted as follows: Model 1—unadjusted analysis; Model 2—adjusted for age, gender, and BMI; Model 3—further adjusted as in Model 2, plus race, education, and marital status; Model 4—adjusted as in Model 3, with additional inclusion of hypertension and diabetes mellitus; Model 5—comprehensive model including all variables from Model 4, plus smoking status, physical activity status, blood urea nitrogen, creatinine, and uric acid. The five models had a progressive relationship; that is, Model 1 remained unadjusted; Model 2 added personalized variables; Model 3 incorporated further demographic variables; Model 4 included clinical variables to account for metabolic and health-related factors that may confound the relationship; Model 5 was the fully adjusted model, which included all the aforementioned predictors, allowing researchers to isolate the effect of GS and rGS on the kidney stone risk. The set of models was based on prior research [8,16].

2.5. Two-sample MR analysis

The MR design is shown in Fig. 1, following the strengthening the reporting of observational studies in epidemiology using MR (STROBE-MR) guideline [17]. Three core assumptions were: (1) genetic instrumental variables (IVs) are strongly associated with the exposure factors in the study; (2) IVs do not have any direct impacts on outcome factors, except through their correlation with exposure factors; (3) IVs are independent of potential confounding factors that may confound the inference results from an MR analysis.

Figure 1.

The Mendelian randomization design following the core assumptions. SNP, single-nucleotide polymorphism.

The IVs for the analysis were selected from the genome-wide association study (GWAS) database provided by the Medical Research Council Integrative Epidemiology Unit at the University of Bristol (https://gwas.mrcieu.ac.uk/). Exposure data were sourced from UK Biobank (ukb-b-10215 and ukb-b-7478), while outcome data were sourced from finn-b-N14_CALCUKIDUR and ebi-a-GCST90018935. The entire population was White European, which avoided the potential bias among different ethnic groups. Since this secondary analysis was conducted using publicly accessible data, ethical approval and informed consent were not required.

The IVs were extracted at a genome-wide significant level (p<5×10⁻⁸). The r² of linkage disequilibrium and the clumping window were set to appropriate levels to avoid a linkage disequilibrium (r²<0.001, 10 000 kb).

The evaluation of IVs encompassed the F-statistic analysis. This study used the R² method to calculate F-statistics.

R² was calculated using the following formula [18]:

$$R^2=2\times \left(1-\text{EAF}\right)\times \text{EAF}\times {\left(\frac{\text{BETA}}{\text{SE}\times \sqrt{N}}\right)}^2$$

The F-statistic for each single-nucleotide polymorphism (SNP) was calculated using the following formula:

$$F=\frac{N-k-1}{k}\times \frac{R^2}{1-R^2}$$

(here, R² represents the proportion of variance in the exposure variable that is accounted for by the SNPs. EAF denotes the effect allele frequency provided by genetic variant datasets. N represents the sample size. BETA denotes the beta value of exposure provided by genetic variants datasets. k is the number of genetic variants.)

Furthermore, the Phenoscanner tool (http://www.phenoscanner.medschl.cam.ac.uk) was used to examine all SNPs extracted from the relevant exposure datasets to ensure independence and adherence to the exclusion criteria. SNPs associated with known confounders, such as obesity and cardiovascular disease, were excluded.

In two-sample MR analyses, the inverse-variance weighted (IVW) method was used as the main statistical model. This method is a meta-analysis of the variants' Wald ratios, weighted by their associated uncertainty. The MR-Egger and weighted median methods were also employed in the analyses. Compared with the MR-Egger and weighted median methods, the IVW method provides the most unbiased estimates in the absence of horizontal pleiotropy, which is the case in this study [19].

The sensitivity analysis included a heterogeneity test and pleiotropy identification. The IVW and MR-Egger methods were used to assess heterogeneity and random-effects IVW methods were used if heterogeneity was detected (p<0.05). The MR-Egger intercept test was used to assess potential pleiotropy. A p-value >0.05 was considered to indicate no pleiotropy.

2.6. Statistical and analysis platforms

The cross-sectional analysis was performed using R software (version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria) and IBM SPSS Statistics (version 27; IBM, Armonk, NY, USA). The MR analysis was performed using the TwoSampleMR package (version 0.6.3) and the MR-PRESSO package (version 1.0) in R software. The detailed information on the packages used can be checked at the following URLs: https://rdrr.io/github/MRCIEU/TwoSampleMR/ and https://rdrr.io/github/rondolab/MR-PRESSO/, respectively. The statistical significance was assessed using a significance threshold of p<0.05.

3. Results

3.1. Participants’ general characteristics

After selection, 9306 participants were included from the NHANES database between 2011 and 2014. There were 302 participants with kidney stones, as well as 9004 participants without kidney stones. Table 1 presents the demographic and clinical characteristics of the study population.

Table 1.

Demographic and clinical characteristics of included participants according to the status of KS.

Characteristic	Total (n=9306)	Without KS (n=9004)	With KS (n=302)	p-Value
Age, year	48.5±17.5	48.1±17.5	58.7±16.4	<0.001
Gender				0.3
Male	4603	4542	161
Female	4703	4462	141
Race				0.12
Non-Hispanic White	3881	3751	130
Non-Hispanic Black	2103	2029	74
Mexican American	1067	1030	37
Other Hispanic	864	832	32
Other	1391	1362	29
Marital status				<0.001
Married	8593	8302	291
Unmarried	713	702	11
Education				<0.001
Less than high school	1957	1860	97
High school or equivalent	2031	1955	76
College or above	5318	5189	129
Body mass index, kg/m2	29.0±7.0	29.0±7.0	30.2±7.3	0.003a
Underweight (<18.5)	156	149	7
Normal (18.5–24.9)	2694	2626	68
Overweight (25.0–29.9)	3006	2915	91
Obese (≥30.0)	3450	3314	136
Hypertension				<0.001
No	5964	5879	85
Yes	3342	3125	217
Diabetes mellitus				<0.001
Yes	1134	1020	114
No	7926	7746	180
Borderline	246	238	8
Smoking status				<0.001
Never	5342	5187	155
Former	2084	1988	96
Current	1880	1829	51
Physical activity status
Vigorous				0.13
Yes	1637	1594	43
No	7669	7410	259
Moderate				0.8
Yes	3107	3004	103
No	6199	6000	199
Grip strength, kg	37.2±11.2	37.3±11.6	32.0±10.1	<0.001
Relative grip strength	1.33±0.48	1.34±0.48	1.12±0.44	<0.001
Blood urea nitrogen, mmol/L	13.1±5.7	12.8±5.0	21.6±14.2	<0.001
Creatinine, mg/dL	0.9±0.5	0.9±0.3	1.8±2.0	<0.001
Uric acid, mg/dL	5.4±1.4	5.4±1.4	6.1±1.8	<0.001

KS, kidney stones.

Note: the data are presented as n or mean±standard deviation.

^a The p-value of 0.003 corresponds to the between-group comparison of mean body mass index (continuous variable). The distribution difference of body mass index categorical groups between groups was p=0.006.

In the whole population, the mean GS of participants without kidney stones was significantly higher than that of participants with kidney stones (37.3 [standard deviation, SD 11.6] kg vs. 32.0 [SD 10.1] kg, p<0.001). A similar result was observed when comparing the mean levels of rGS between participants with and without kidney stones (1.12 [SD 0.44] vs. 1.34 [SD 0.48], p<0.001). The GS and rGS levels are shown in Fig. 2. In addition, chi-square analyses demonstrated differences in age, marital status, education, BMI, hypertension, diabetes mellitus, and smoking status, between participants with and without kidney stones. Statistical differences were also observed in blood urea nitrogen, creatinine, and uric acid between the two groups. No statistical differences were found in gender, race, or physical activity status between the two groups (Table 1).

Figure 2.

GS and rGS levels in participants without and with KS. (A) GS; (B) rGS. GS, grip strength; rGS, relative grip strength; KS, kidney stones. ∗∗∗p<0.001.

3.2. Factors associated with rGS and GS

Table 2 displays the potential influencing factors among four groups divided based on the quartiles of rGS distribution (Q1–Q4). Clear trends could be observed from Q1 to Q4: age declined; the number of participants with kidney stones declined (from 135 in Q1 to 42 in Q4). There were statistically significant differences in race, gender, marital status, education, BMI, hypertension, diabetes mellitus, smoking status, physical activity status, blood urea nitrogen, creatinine, and uric acid among the four groups participants with kidney stones (p<0.001 for all variables).

Table 2.

Demographic and clinical characteristics of included participants according to relative grip strength.

Characteristic	Relative grip strength				p-Value
	Q1 (n=2327)	Q2 (n=2327)	Q3 (n=2326)	Q4 (n=2326)
Age, year	56.4±17.2	49.2±17.5	47.2±17.0	41.1±14.8	<0.001
Gender					<0.001
Male	240	679	1490	2194
Female	2087	1648	836	132
Race					<0.001
Non-Hispanic White	998	957	943	983
Non-Hispanic Black	507	532	481	583
Mexican American	337	268	272	190
Other Hispanic	272	225	217	150
Other	213	345	413	420
Marital status					<0.001
Married	2198	2169	2143	2083
Unmarried	129	158	183	243
Education					<0.001
Less than high school	638	464	436	423
High school or equivalent	568	457	478	527
College or above	1121	1406	1412	1376
Body mass index, kg/m2	34.5±8.2	29.0±6.1	27.2±5.4	25.3±4.0	<0.001a
Underweight (<18.5)	11	16	55	74
Normal (18.5–24.9)	195	610	822	1067
Overweight (25.0–29.9)	499	846	765	896
Obese (≥30.0)	1622	855	684	289
Hypertension					<0.001
No	1077	1505	1599	1783
Yes	1250	822	727	543
Diabetes mellitus					<0.001
Yes	518	281	234	101
No	1719	1987	2022	2198
Borderline	90	59	70	27
Smoking status					<0.001
Never	1463	1439	1267	1173
Former	530	492	598	464
Current	334	396	461	689
Physical activity status
Vigorous					<0.001
Yes	192	318	464	663
No	2135	2009	1862	1663
Moderate					<0.001
Yes	625	740	815	927
No	1702	1587	1511	1399
Kidney stone					<0.001
Yes	135	73	52	42
No	2192	2254	2274	2284
Blood urea nitrogen, mmol/L	14.2±7.3	12.5±5.8	13.0±4.9	12.7±4.2	<0.001
Creatinine, mg/dL	0.89±0.65	0.85±0.54	0.92±0.31	0.99±0.46	<0.001
Uric acid, mg/dL	5.44±1.49	5.13±1.41	5.47±1.48	5.74±1.27	<0.001

Note: the data are presented as n or mean±standard deviation. Q1–Q4 refer to the first to fourth quartiles of relative grip strength.

^aThe between-group comparison of the mean body mass index (continuous variable) and the difference in the distribution of categorical body mass index groups both have p-values <0.001.

Similar results were obtained in analyses of different variables across the four GS groups, as presented in Supplementary Table 1.

3.3. Relationship of GS or rGS with nephrolithiasis

3.3.1. GS levels and the risk of nephrolithiasis

In our analysis, we explored the relationship between GS levels and the risk of kidney stones. Using five logistic regression models based on the quartile distribution of GS (Q1–Q4), we observed a decreasing trend in the prevalence of kidney stones as GS levels increased. This trend was visually represented in Table 3.

Table 4.

Associations between quartile groups of GS, rGS, and kidney stones.

Characteristic	Model 1		Model 2		Model 3		Model 4		Model 5
	OR (95% CI)	p-Value	OR (95% CI)	p-Value	OR (95% CI)	p-Value	OR (95% CI)	p-Value	OR (95% CI)	p-Value
GS, kg
Q1	Reference		Reference		Reference		Reference		Reference
Q2	0.51 (0.37–0.69)	<0.001	0.56 (0.41–0.76)	<0.001	0.59 (0.43–0.80)	<0.001	0.70 (0.51–0.96)	0.025	0.76 (0.54–1.08)	0.13
Q3	0.57 (0.43–0.77)	<0.001	0.62 (0.47–0.84)	0.002	0.64 (0.48–0.87)	0.004	0.75 (0.55–1.02)	0.064	0.66 (0.46–0.93)	0.019
Q4	0.24 (0.16–0.35)	<0.001	0.29 (0.19–0.43)	<0.001	0.30 (0.20–0.45)	<0.001	0.38 (0.25–0.57)	<0.001	0.32 (0.20–0.51)	<0.001
rGS
Q1	Reference		Reference		Reference		Reference		Reference
Q2	0.53 (0.39–0.70)	<0.001	0.57 (0.42–0.76)	<0.001	0.60 (0.44–0.82)	0.001	0.71 (0.52–0.97)	0.030	0.71 (0.50–1.01)	0.053
Q3	0.37 (0.27–0.51)	<0.001	0.41 (0.29–0.57)	<0.001	0.43 (0.30–0.61)	<0.001	0.53 (0.37–0.75)	<0.001	0.48 (0.32–0.71)	<0.001
Q4	0.30 (0.21–0.42)	<0.001	0.36 (0.24–0.53)	<0.001	0.37 (0.25–0.55)	<0.001	0.51 (0.34–0.77)	0.001	0.45 (0.28–0.71)	<0.001

OR, odds ratio; CI, confidence interval; GS, grip strength; rGS, relative grip strength.

The predictive accuracy of the model for GS was evaluated using ROC curves, which were depicted in Fig. 3. The area under the curve for GS was 0.812, indicating a high level of accuracy in assessing the risk of kidney stone formation associated with GS.

Figure 3.

ROC curves for evaluating the diagnostic efficacy of the model. (A) Grip strength; (B) Relative grip strength. ROC, receiver operating characteristic; AUC, area under the curve.

Associations between quartile groups of GS and kidney stones are shown in Table 4. In the most comprehensive model, Model 5, the OR for the highest quartile (Q4) compared with the lowest quartile (Q1) for GS was 0.32, with a 95% CI of 0.20–0.51, and the result was statistically significant (p<0.001). This suggests that individuals in the highest quartile of GS have a significantly lower risk of developing kidney stones compared with those in the lowest quartile.

Table 3.

Logistic regression analysis of the association between GS, rGS and kidney stone risk.

Trait and modela	OR (95% CI)	p-Value
GS
Model 1	0.96 (0.95–0.97)	<0.001
Model 2	0.97 (0.96–0.98)	<0.001
Model 3	0.97 (0.96–0.99)	<0.001
Model 4	0.98 (0.96–0.99)	<0.001
Model 5	0.96 (0.95–0.98)	<0.001
rGS
Model 1	0.33 (0.25–0.43)	<0.001
Model 2	0.50 (0.36–0.70)	<0.001
Model 3	0.51 (0.36–0.72)	<0.001
Model 4	0.54 (0.39–0.77)	<0.001
Model 5	0.39 (0.26–0.58)	<0.001

GS, grip strength; rGS, relative grip strength; OR, odds ratio; CI, confidence interval.

^aModels 1–5 represent sequential adjustment for confounding factors: Model 1—unadjusted analysis; Model 2—adjusted for age, gender, and BMI; Model 3—adjusted as in Model 2, plus race, education, and marital status; Model 4—adjusted as in Model 3, plus hypertension and diabetes mellitus; Model 5—fully adjusted model including all variables from Model 4, plus smoking status, physical activity status, blood urea nitrogen, creatinine, and uric acid.

3.3.2. rGS levels and risk of nephrolithiasis

Similarly, we examined the association between rGS levels and the risk of kidney stones. The same five logistic regression models were applied to the quartile distribution of rGS (Q1–Q4), and we also noted a decreasing trend in the prevalence of kidney stones as rGS levels increased, as shown in Table 3.

The accuracy of the model for rGS was also assessed via ROC curves, as presented in Fig. 3. The area under the curve for rGS was 0.808, which similarly indicates a high degree of accuracy in assessing the risk of kidney stone formation associated with rGS.

Associations between the quartile division of rGS and kidney stones are shown in Table 4. For rGS, in Model 5, the OR for Q4 versus Q1 was 0.45, with a 95% CI of 0.28–0.71, and this finding was also statistically significant (p<0.001). This implies that individuals with higher rGS levels (in the top quartile) have a notably reduced risk of kidney stones when compared with those with the lowest rGS levels.

3.4. MR results

The retrospective analysis based on the NHANES database provided evidence supporting the protective effect of GS on nephrolithiasis. The MR analysis further strengthened these findings by providing causal inferences.

The F-statistic of GS (right) was 24.49, and the F-statistic of GS (left) was 15.56 (Supplementary Table 2). The analysis demonstrated robust F-statistics exceeding 10, indicating strong associations with the exposure variables.

The sensitivity analyses revealed the presence of heterogeneity (p<0.05). To cope with the heterogeneity, this study utilized the random-effects IVW method to calculate the associations. No evidence of horizontal pleiotropy was revealed under an MR-Egger test (Supplementary Table 2).

The results of MR analyses for GS on kidney stones are presented in Fig. 4. Right-hand GS was proven to be a protective factor against kidney stones (IVW method, dataset finn-b-N14_CALCUKIDUR: OR 0.64, p=0.016; dataset ebi-a-GCST90018935: OR 0.79, p=0.037).

Figure 4.

Associations of genetically predicted grip strength with kidney stones. GWAS, genome-wide association study; OR, odds ratio; MR, Mendelian randomization; CI, confidence interval; FinnGen, finn-b-N14-CALCUKIDUR; EBI, ebi-a-GCST90018935.

4. Discussion

In this study, we investigated the impacts of GS and rGS on the risk of kidney stones using data from 9306 participants in the NHANES database. The findings revealed negative correlations between GS and kidney stones and between rGS and kidney stones. Furthermore, a two-sample MR study was implemented to explore the causal relationship between GS and kidney stones, revealing a consistent inverse causal association between right-hand GS and kidney stones across various datasets. These results partially confirmed the potential preventive role of GS, and by extension, muscle strength, in the process of lithogenesis.

The analysis of the NHANES database revealed that both GS and rGS were negatively associated with the prevalence of kidney stones. Notably, differences in baseline characteristics between participants with and without kidney stones were observed. To account for these differences, covariates were incorporated into logistic regression models using various specifications, and the results remained consistent. When GS and rGS were divided into quartiles, a trend of decreasing kidney stone prevalence with increasing GS was evident. This study is exploratory in nature and that the models are intended to assess associations rather than predict outcomes.

A two-sample MR study was conducted to further explore the underlying relationship. To avoid the limitations of a single dataset, two distinct GWAS datasets were selected as sources of IVs for kidney stones, adhering strictly to the STROBE-MR guidelines. The results showed a negative causal relationship between right-hand GS and kidney stones, whereas left-hand GS was not associated with kidney stones. This intriguing distinction prompted a re-examination of the exposure definitions in the NHANES study. In the NHANES study, GS was measured three times for each hand, and the maximum value out of six measurements was used as the standard. Given that the majority of the population are right-handed (up to 90%) [20], the maximum GS often equaled right-hand GS, while left-hand GS was less representative and potentially overlooked as a measure of true muscle strength. In summary, the MR analysis successfully validated a significant negative causal relationship between GS and kidney stones, with heterogeneity corrected and no evidence of horizontal pleiotropy. This strengthens the argument for the preventive role of GS in kidney stone disease.

The formation of kidney stones within the body is a complicated physicochemical process involving nucleation, crystal attachment, supersaturation, and ultimately precipitation into solid masses [21]. This intricate mechanism underscores the multifaceted risk factors associated with kidney stone development, which span demographic variables such as sex, age, and ethnicity, alongside dietary habits [[22], [23], [24], [25]]. High sodium intake, excessive protein consumption, frequent sugary drink consumption, and inadequate hydration are dietary factors known to elevate the risk of stone formation [26,27]. Additionally, obesity and metabolic disorders contribute significantly to the increased likelihood of developing kidney stones, potentially through mechanisms such as altered metabolism, insulin resistance, and modifications in urinary pH [28,29]. The adipose tissue correlated with obesity also contributes to oxidative stress and inflammation, which induce the formation of Randall's plaque [6]. Exercise habits have also been validated to be related to the etiology of nephrolithiasis [16]. Apart from lifestyle factors, the influencing factors also include genetic abnormalities (such as hypercalciuria and cystinuria) and environmental factors (such as high temperature and low humidity) [21,30]. Systemic conditions like hypertension, diabetes mellitus, and gastrointestinal disorders also contribute to the risk by altering urine composition, renal function, and electrolyte balance [21].

Muscle strength, particularly as measured by GS, exhibits an inverse relationship with systemic inflammation and oxidative stress [31,32]. Low muscle mass, strength, and sarcopenia are related to insulin resistance, which can increase the risk of hypercalciuria and low urinary pH, key factors in kidney stone formation [[28], [29], [30], [31], [32], [33]]. In addition, sarcopenia was found to be associated with inflammatory cytokines [10,34], which provides a potential explanation for the mechanism of nephrolithogenesis. One randomized clinical trial focused on muscular training and its effects on kidney function in patients following kidney transplant; the results revealed that muscular training may improve kidney function, providing a valuable reference for our research [35]. However, further research is needed to investigate the potential protective effects of muscular training specifically in patients with kidney stones.

This research has several limitations that should be considered when interpreting the findings. First, the cross-sectional study design inherently limits our ability to establish causal relationships and may introduce biases, particularly from self-reported data. Second, important risk factors, such as daily water intake and dietary habits, were not included in the analysis, which could impact the accuracy of the results. Finally, the study did not differentiate among types of kidney stones, which may affect the generalizability of the conclusions, as related risk factors are highly associated with uric acid stones and oxalate stones, while struvite stones are more related to urinary infection [36]. These limitations highlight the need for further research with more comprehensive data to validate the findings.

5. Conclusion

In summary, this study, by utilizing the NHANES database in conjunction with MR methods, has revealed for the first time a potential association between GS and kidney stones, offering new insights for future research. Based on these findings, further research could develop predictive models for kidney stone incidence, incorporating variables such as GS and rGS. Future research should consider adopting prospective cohort designs to overcome the limitations of the current study design, while also delving into the biological mechanisms underlying the relationship between GS and kidney stones, aiming to provide a more robust scientific foundation for the prevention and treatment of kidney stones.

Author contributions

Study concept and design: Guangyuan Zhang.

Data acquisition: Xiangyu Zhang.

Data analysis: Xiangyu Zhang, Haiyan Ju.

Drafting of the manuscript: Xiangyu Zhang.

Critical revision of the manuscript: Bo Xiao.

Conflicts of interest

The authors declare no conflict of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article.