Background

Gallstones (GS), common digestive diseases worldwide, are solid substances that form in the gallbladder or bile ducts, typically composed of cholesterol, bile salts, or other substances1. Epidemiological investigations have found that the prevalence of GS varies widely in different regions1. For example, the prevalence of GS is significantly higher in developed regions, such as the United States and Europe, than in Asia2. The annual healthcare cost associated with GS in the United States is approximately $6 billion3. Thus, GS places a heavy burden on public health and healthcare. Most patients with GS are asymptomatic, and a small percentage of patients may present with clinical symptoms, mainly in the form of sudden onset of epigastric or right upper abdominal pain4. Despite being benign, GS can cause several complications, such as acute cholecystitis, choledocholithiasis, acute cholangitis, and acute pancreatitis5. Additionally, GS is a significant risk factor for gallbladder cancer1. GS is usually treated surgically, with laparoscopic cholecystectomy being the preferred treatment option5.

The pathogenesis of GS is complex, with numerous risk factors6,7. Among these, the role of obesity, particularly abdominal obesity, in stone formation has been the focus of GS etiology8,9,10. Body mass index (BMI) is the most commonly used obesity indicator. Previous studies have found that for every 5-unit increase in BMI, the incidence of GS increases by 63%10. However, BMI has significant limitations, as it does not reflect body fat distribution or differences in adiposity between sexes, making it a rough indicator of overweight rather than an accurate measure of body fat. In this context, Relative fat mass (RFM) has emerged as a novel indicator for assessing obesity, particularly abdominal obesity.

RFM is a recently proposed indicator of obesity (particularly abdominal obesity) in recent years11. RFM is an indicator based on the ratio of height to waist circumference, adjusted by sex, and effectively reflects both fat distribution and sex differences in fat content11. Thus, RFM not only considers the distribution of fat (height and waist circumference ratio), but also distinguishes differences in fat distribution and content by sex, which makes RFM more sensitive and accurate in the assessment of obesity and obesity-related diseases. Previous studies have shown that RFM is closely associated with obesity-related diseases, such as non-alcoholic fatty liver disease, metabolic syndrome, and cardiovascular disease12,13,14,15. However, the relationship between the RFM and GS remains unclear.

To address this gap, we hypothesize that RFM may be associated with an increased prevalence of GS and could be a valuable predictor for the age at first GS surgery. Therefore, the aim of our study was to investigate the association between RFM and GS prevalence in adults, assess its predictive value for GS risk, and explore its potential impact on the timing of surgical intervention for GS.

Methods

Source of participants

The participants were drawn from the National Health and Nutrition Examination Survey (NHANES), a survey program administered by the National Center for Health Statistics. The program is a valuable resource for understanding and tracking health trends and needs in the United States. The NHANES plays a crucial role in identifying health priorities, developing disease prevention strategies and public health policies, preventing chronic diseases, improving nutrition, and promoting healthy lifestyles. It uses a stratified multi-stage probability sampling approach (consisting of the following steps: stratification, primary sampling units, secondary sampling units, household sampling, and oversampling) to ensure that the sample is nationally representative. The NHANES consists of four main components: demographic, screening, laboratory, and questionnaire data. The NHANES is updated every 2 years and covers a population of approximately 10,000 people. The program was suspended in March 2020 owing to the COVID-19 pandemic. Since 2017, NHANES has collected information about GS in the population through questionnaires; therefore, we selected participants from this period (2017– March 2020) as the study population. Overall, 15,560 participants were included in the study period, of which 6328 were younger than 20 years of age and were excluded. Participants who lacked information on the GS (22) or RFM (1168) were excluded. In addition, 64 individuals were excluded due to a lack of information on covariates such as hypertension (10 individuals), diabetes (3 individuals), coronary heart disease (25 individuals), asthma (5 individuals), cancer (4 individuals), education (9 individuals), marital status (5 individuals), and smoking status (3 individuals). Finally, 7978 individuals were included in the study. Figure 1 shows the details of the selection criteria for the study population.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Flowchart for participants from NHANES.

The data from the NHANES used in our study is cross-sectional in nature, and participants were not followed up continuously during the study period. The study period spans from 2017 to March 2020, and we selected data collected during this period for our analysis. This cross-sectional design enables us to explore the association between RFM and the risk of GS at a specific point in time, rather than over a longitudinal follow-up period.

Data collection

GS was assessed based on a questionnaire that focused on the question, “Have you ever been told that you have GS?” and “Age at the time of your first GS surgery.” The questionnaire was only administered to adults aged ≥ 20 years; therefore, patients younger than 20 years were excluded from this study. Multiple covariates (demographic, examination, and laboratory data) were included in this study, with details provided in Supplementary Table 1. All participants underwent two 24-hour dietary recalls (these recalls were conducted on non-consecutive days within the same survey cycle); therefore, the average consumption of the two recalls was used in the analysis. The measurement procedures for all covariates are available on the NHANES website (www.cdc.gov/nchs/nhanes/). RFM was calculated as follows: RFM = 64-(20 × height/waist circumference)+(12 ×sex); sex equals 0 for men and 1 for women16.

Ethics statement

All participants signed an informed consent form; therefore, the NHANES was exempt from additional licensing or ethical review.

Statistical methods

Categorical variables were expressed as numbers and percentages (%), and comparisons between groups were made using Pearson’s χ2 test or Fisher’s exact test. Continuous variables are expressed as medians and interquartile ranges, and comparisons of continuous variables between groups were performed using a t-test or one-way ANOVA. Several multivariate logistic regression models were constructed to investigate the relationship between the RFM and GS prevalence. Multiple linear regression analyses were used to explore the relationship between RFM and patient age at the time of the first GS surgery. Model 1 was unadjusted, and Model 2 was adjusted for age, sex, race, and marital status. Model 3 was additionally adjusted for education level, physical activity, smoking status, alcohol consumption, diabetes mellitus, hypertension, asthma, cancer, coronary heart disease, total cholesterol, triglycerides, blood glucose, blood pressure, total cholesterol, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, alanine aminotransferase, aspartate aminotransferase, hemoglobin A1c, ferritin, poverty income ratio, total calories, total fat, and total sugar. To further investigate the relationship between RFM, GS prevalence, and age at which patients first underwent GS surgery, we used a generalized additive model with smooth curve fitting (penalized spline method). The use of generalized additive model enabled us to explore potential thresholds or non-linear trends that could have been overlooked using simpler models. For nonlinear relationships, we used the logarithmic likelihood ratio test to calculate inflection points. Subgroup analyses were performed to assess the relationship between RFM and GS in various populations, including age, sex, race, hypertension, and diabetes. Receiver operating characteristic (ROC) curves were used to analyze the predictive ability of RFM for GS development. Statistical significance was defined as P < 0.05. All analyses were performed using the R software version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria), details are presented in the Supplementary material.

Results

Characteristics of participants

Overall, 7978 participants were finally included in this study, comprising 3900 males and 4078 females. Of these, 828 participants had a history of GS. RFM was significantly higher in the GS group compared with the non-GS group (43.80 [35.09–47.78] vs. 34.91 [29.31–42.72], P < 0.001). In the GS group, non-Hispanic white individuals were the most numerous (42.27%), followed by Mexican Americans (24.28%). Participants in the GS group were more likely to have hypertension, diabetes, asthma, coronary heart disease, and cancer than those in the non-GS group (All P < 0.001). In addition, the serum triglyceride and hemoglobin A1c levels were significantly higher in the GS group (All P < 0.001). Further details are provided in Table 1.

Table 1 Baseline characteristics of participants.
Full size table

Logistic regression results between RFM and prevalence of GS

We observed a positive correlation between RFM and GS prevalence (Table 2). In Model 3, after adjusting for potential confounders, we found that for every 1-unit increase in RFM, GS prevalence increased by 9% (odds ratio (OR): 1.09, 95% confidence interval (CI): 1.07, 1.11). After converting RFM from a continuous variable to a categorical variable (tertiles) for sensitivity analyses, tertile 3 was associated with a significant 3.11-fold increase in GS prevalence (OR = 3.11, 95% CI: 2.20, 4.38) compared with the lowest RFM tertile (tertile 1).

Table 2 Results of logistic regression analysis between RFM and prevalence of gallstone.
Full size table

Dose-response relationship and threshold effect analysis of RFM and GS prevalence

The relationship between the RFM and GS prevalence was further explored using additive generalized modeling and smoothed curve fitting. Figure 2; Table 3 show a positive (nonlinear) correlation between RFM and GS prevalence. Considering the effect of the saturation threshold between them, the likelihood natural ratio test determined the optimal RFM threshold to be 40.99.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Density dose-response relationship between RFM with gallstones prevalence. The area between the upper and lower dashed lines is represented as 95% CI. Each point shows the magnitude of the RFM and is connected to form a continuous line. Adjusted for all covariates except effect modifier.

Table 3 Two-piecewise linear regression and logarithmic likelihood ratio test explained the threshold effect analysis of RFM with gallstone prevalence.
Full size table

Subgroup analysis

To assess whether the relationship between RFM and GS prevalence holds in different populations, we performed additional subgroup analyses. The results in Table 4 suggest that the positive correlation between RFM and GS prevalence remained stable in all populations, although there were subtle differences between populations. For example, for every 1-unit increase in RFM, the prevalence of GS in men would increase by 6% (OR = 1.06, 95% CI: 1.02, 1.09), whereas the prevalence of GS in women would increase by 11% (OR = 1.11, 95% CI: 1.09, 1.13). For patients with diabetes, each 1-unit increase in RFM was associated with a 6% increase in GS prevalence (OR = 1.06, 95% CI: 1.02, 1.11) and a 10% increase in the non-diabetic population (OR = 1.10, 95% CI: 1.08, 1.12). Additional details are presented in Table 4.

Table 4 Subgroup analysis to evaluate the relationship between RFM and gallstone prevalence in different populations.
Full size table

ROC curve analysis

We plotted the ROC curve of the RFM to assess its predictive value for GS prevalence (Fig. 3). It was found that RFM had a higher predictive value for the risk of GS prevalence (69.66% [95% CI: 67.78–71.53]) compared to BMI (63.96% [95% CI: 61.99–65.92]) and waist circumference (63.99% [95% CI: 62.08–65.90]). Figure 3 pairs of information are presented in Table 5.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

ROC curves for predicting the risk of gallstones.

Table 5 Information of ROC curves in Fig. 3.
Full size table

Results of multivariate linear analysis between RFM and patients’ age at first GS surgery

In Model 1, for every 1-unit increase in RFM, the patient’s age at first GS surgery would be 0.45 years earlier (β = −0.45, 95% CI: −0.59, −0.31). When we adjusted for potential confounders, for each 1-unit increase in RFM, patients might be 0.19 years earlier at the time of their first GS surgery (β=−0.19, 95% CI: −0.40, 0.02), though this result did not reach statistical significance (Table 6).

Table 6 Relationship between RFM and patient’s age at first gallstone surgery.
Full size table

Dose-response relationship and threshold effect analysis of RFM and patient’s age at first GS surgery

To further investigate the relationship between RFM and patient age at first GS surgery, generalized additive modeling and smoothed curve fitting were performed. The results suggested a nonlinear relationship between RFM and patient age at the time of the first GS surgery (Fig. 4). The results in Table 7 demonstrated that the threshold for the effect of RFM on patient age at the first GS surgery was 30.99.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Density dose–response relationship between RFM and age at first gallstone surgery. The area between the upper and lower dashed lines is represented as the 95% CI. Each point shows the magnitude of RFM and is connected to form a continuous line. Adjusted for all covariates except the effect modifier.

Table 7 Two-piecewise linear regression and logarithmic likelihood ratio test explained the threshold effect analysis of RFM with patient’s age at first gallstone surgery.
Full size table

Discussion

This is the first study to elucidate the relationship between RFM and GS prevalence in adults, and patient age at the time of first GS surgery. The results showed that RFM was positively associated with GS prevalence. When RFM was converted from a continuous to a categorical variable, GS prevalence increased significantly as RFM increased. A positive correlation between RFM and GS prevalence was established in all the patient groups analyzed in this study. In addition, we observed a nonlinear relationship between RFM and the age at first GS surgery and found a significant threshold effect of RFM. Specifically, when RFM reaches 30.99, the age at first GS surgery is significantly younger, suggesting that a higher RFM may be associated with earlier surgical intervention. Moreover, the predictive value of RFM for GS prevalence was significantly higher than those of BMI and waist circumference. Therefore, this study provides strong evidence that RFM can be used as a predictive tool for identifying individuals at high risk of GS.

The pathogenesis of GS is complex and involves several factors. For example, mutations in certain genes (e.g., ABCG5/ABCG8) increase the risk of GS, leading to a certain familial predisposition to the condition17. The risk of developing GS varies across countries and regions. Generally, the incidence of GS is significantly higher in developed regions, such as Europe and the United States, than in Asia18. In addition, the risk of GS in women is usually higher than that in men4,5, which is mainly influenced by high estrogen levels in women. In this study, the prevalence of GS was higher in women than in men (71.14% vs. 28.86%; P< 0.001). Estrogen not only increases HMG-CoA reductase activity in the liver, promoting cholesterol synthesis and secretion19,20but also inhibits the synthesis of bile acids21, leading to the formation of cholesterol crystals and stones in supersaturated bile. Age is also an important factor in the development of GS. In this study, the age was significantly higher in the GS group than in the non-GS group (60 [46–70] vs. 50 [34–63], P< 0.001). This was consistent with the results of previous studies22,23. In old age, gallbladder motility is weakened, and emptying ability decreases, leading to cholestasis24. The composition of bile changes with age, particularly when the ratio of phospholipids to bile acids decreases, leading to decreased cholesterol solubility and increased risk of cholesterol crystallization22,23.

Obesity, particularly abdominal obesity, is significantly associated with an increased incidence of GS25,26. Therefore, the role of obesity in the stone formation has been the focus of GS etiology research8,9,10. However, the specific mechanisms underlying obesity-induced GS formation have not yet been fully elucidated. Existing studies suggest that obesity can promote stone formation by affecting several aspects, such as cholesterol synthesis and secretion, bile composition, gallbladder function, and inflammation.

Abnormal cholesterol metabolism in patients with obesity is a key factor in GS formation. Studies have shown that increased hepatic synthesis of cholesterol in obese individuals leads to the overproduction of cholesterol in bile, which puts bile in a state of cholesterol supersaturation25,26. When cholesterol is supersaturated in bile, it tends to form cholesterol crystals, which is an early process in GS formation25,26. Additionally, a decreased bile acid-to-lecithin ratio is associated with obesity. Bile acids and lecithin play key roles in maintaining cholesterol in a soluble state, and reductions in both further promote the precipitation of cholesterol crystals27. In addition, increased fasting gallbladder volume and abnormal postprandial gallbladder emptying in patients with obesity lead to cholestasis, which may be associated with the abnormal secretion of cholecystokinin (CCK)28, a phenomenon that was confirmed in mice with a knockout of the CCK gene29. Microsomal triglyceride transfer proteins (MTTP) play important roles in lipid metabolism by acting as lipid transport proteins. Thus, MTTP mutations increase the risk of stone formation by affecting cholesterol transport and reabsorption in the intestine30. Moreover, activation of the ileal Fxr-Fgf15 pathway in an obese rat model inhibits cholesterol 7α-hydroxylase (CYP7A1) expression, which affects endogenous cholesterol conversion and disrupts the homeostasis of bile constituents and increases the potential for stone formation31. Therefore, obesity can induce and promote stone formation by affecting bile synthesis, excretion, and reabsorption.

In addition, patients with obesity are often accompanied by insulin resistance and metabolic syndrome, which are important risk factors for GS formation32,33,34. Insulin resistance increases hepatic cholesterol synthesis and decreases bile acid synthesis, further contributing to cholesterol supersaturation32. Metabolic syndrome is also characterized by hypertriglyceridemia and low high-density lipoproteinemia, which are not only related to bile composition, but may also increase the risk of stone formation by altering gallbladder function (e.g., delayed gallbladder emptying)35. Obesity is also closely associated with a systemic low-grade chronic inflammatory state, particularly in patients with obesity, where adipose tissue overproduces pro-inflammatory factors (e.g., TNF-α, IL-6), which not only trigger systemic metabolic disorders, but also have a direct effect on the gallbladder, resulting in chronic inflammation and fibrosis in the gallbladder wall36,37. These inflammatory responses may further weaken the contractile function of the gallbladder, increasing the risk of bile retention38,39. Recent studies have also shown that obesity may indirectly contribute to GS formation by affecting the gut flora40. Individuals with obesity experience significant changes in the gut flora, particularly an increase in bile salt-hydrolyzing bacteria, which break down bile salts and reduce the reabsorption of bile acids in the gut, thereby affecting the circulation and concentration of bile acids, and further exacerbating cholesterol supersaturation in the bile41,42. Additionally, intestinal dysbiosis may affect bile secretion via the enterohepatic axis, increasing the risk of stone formation41,43. In conclusion, obesity contributes to the development of GS through a combination of mechanisms, and an in-depth understanding of these mechanisms could contribute to better prevention and treatment of obesity-associated GS.

RFM, an obesity indicator that integrates waist circumference, height, and sex, effectively reflect the distribution of adipose tissue and obesity (particularly abdominal obesity)11. BMI is the most commonly used indicator for evaluating obesity; however, the calculation of BMI is based only on weight and height and does not consider the distribution of adipose tissue in the body. Therefore, BMI can only roughly reflect whether the weight is overweight, while RFM is calculated as the ratio of height to waist circumference, which can more directly reflect the body fat content, making RFM have higher sensitivity and accuracy in the assessment of obesity. In addition, BMI does not differentiate between the sexes; therefore, BMI may have some limitations in the assessment of obesity in different sexes. Gender adjustment of the RFM makes it more applicable and accurate for different sexes. Abdominal fat (particularly visceral fat) is closely associated with various metabolic diseases, and BMI is only a total body mass index that does not reflect the local fat distribution. RFM can be used to assess abdominal obesity more accurately by introducing waist circumference into the formula. Recent studies have found that RFM is closely associated with obesity-related diseases, such as non-alcoholic fatty liver disease, metabolic syndrome, and cardiovascular disease12,13,14,15. Our study also found that RFM was closely associated with GS, confirming the significance of RFM in predicting the risk of GS.

Our study had several strengths. First, we elucidated for the first time in the general adult population the association of RFM with the prevalence of GS and the age of patients at the time of their first GS surgery. Second, the design of stratified multistage probability sampling ensured the diversity and representativeness of the samples. Therefore, our findings reflect the general health status of adults. Third, we adjusted for confounders that could potentially affect the results to ensure that our conclusions are applicable to a more generalized population. Finally, our study confirmed the superior predictive value of RFM over BMI for the risk of developing GS.

However, this study has some limitations. First, because this was a cross-sectional study, we were unable to determine a causal relationship between RFM and GS. Second, the diagnosis of GS and the patient’s age at the time of the first GS surgery were determined using a questionnaire; thus, recall (recollection) bias may be present. Finally, although we adjusted for many potential confounders, we were unable to eliminate the effects of other unmeasured confounders. Despite these limitations, the findings of this study are significant. As a novel validated surrogate for obesity, RFM provides a novel method to assess the risk of GS prevalence.

Based on our findings, it is important to highlight the potential clinical applications of RFM as an adjunctive tool for screening individuals at high risk for GS. Given that RFM has shown a significant association with GS prevalence, it could complement traditional measures such as BMI and waist circumference in identifying high-risk populations. Moreover, the global applicability of RFM as a preventive tool warrants further investigation. Future studies should aim to validate its effectiveness across diverse populations with varying characteristics, and explore additional factors, such as genetic susceptibility and dietary habits, which may influence the relationship between RFM and GS risk. Longitudinal studies are also essential to establish causal relationships and better understand the temporal dynamics of RFM in the development of GS.

Conclusion

In conclusion, this study is the first to examine the relationship between RFM and GS prevalence and patient age at the time of the first GS surgery and to explore the value of RFM in predicting the risk of developing GS. In the future, a multicenter prospective cohort study is necessary to confirm the potential of RFM as an independent predictor of GS occurrence.