Liver Function-Related Indicators and Risk of Gallstone Diseases—A Multicenter Study and a Systematic Review and Meta-Analysis

Purpose of the study: We aim to examine the association between liver function-related indicators and gallstone disease (GSD) risk. Study design: The subjects who participated in the China Multicenter Physical Examination Cohort (CMPEC) were enrolled. Relative odds ratios (ORs) with 95% CIs and standardized mean differences (SMDs) were applied to investigate the effect of liver function-related indicators and GSD risk. Moreover, a systematic review and meta-analysis were conducted until July 2021. Additionally, the results in the CMPEC and the systematic review and meta-analysis were combined by meta-analysis. Finally, the results were validated by a cohort study of the UK Biobank (UKB). Results and conclusions: Totally, 369,931 subjects in CMPEC were included in the study. A total of 28 publications were incorporated into the systematic review and meta-analysis. The pooled analysis suggested that aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), total protein (TP), and low albumin (ALB) were positively associated with the risk of GSD. Meanwhile, GSD present to have higher AST, ALT, gamma-glutamyl transferase (GGT), total bilirubin (TBil), globulin (G), and ALP levels and relatively lower TP and ALB levels than the healthy participants. These results were consistent when stratified by the study design, geographic background, and study quality. Only the association between ALP and GSD risk was validated in the UKB cohort. This study suggests liver function indicators were associated with GSD risk. The results may provide the basis for exploring the etiology of GSD and may help clinicians identify high-risk subjects. Trial Registration: PROSPERO (CRD42020179076).


Introduction
Gallstone disease (GSD) is a common digestive disease.The frequency of GSD varies widely between countries due to differences in genetics, environment, lifestyle, etc.In Europe, about 20% of adults harbor GSD [1].In comparison, the prevalence of GSD in Chinese adults is about 3%-11% [2].The prevalence of GSD in women is higher than in men [3].
During the lifetime of GSD patients, more than a fifth will experience biliary symptoms or complications and require surgical management [4].It is reported that over 50,000 cholecystectomies are performed each year in the United Kingdom.While in the United States, cholecystectomies reach approximately 800,000 and consume nearly 6.0 billion dollars annually, which inflicts heavy health burdens and economic costs [5,6].Furthermore, it is reported that GSD will increase the risk of diabetes [7], tumor [8][9][10], and all-cause mortality [11,12], which will bring severe disease and economic burden to patients.Thus, it is essential to study the cause of GSD for disease prevention and control.
Previous studies suggested liver function indicators were associated with GSD risk [3,13,14].However, due to the relatively smaller sample size, the study design variations, and the geographic background, the reported results were inconsistent [15][16][17].
China Multicenter Physical Examination Cohort (CMPEC) is a multicenter study that enrolled the subjects who participated in the health examination in nine hospitals in Tianjin, Beijing, Chongqing, Sichuan, and Shandong provinces of China between 2015 and 2020, which incorporated about 1.2 million participants [18].
Therefore, we first conduct a cross-sectional study through CMPEC to find the association between liver function-related indicators and GSD risk.Secondly, we conduct a systematic review and meta-analysis aiming to confirm the association.And then, combined the abovementioned results by meta-analysis to investigate the associations between liver function-related indicators and the GSD risk.Ultimately, the present study tried to validate these associations using the cohort study of the UK Biobank (UKB) [19].

Physical Examination.
The physical examination of the subjects in CMPEC was administered and recorded by trained investigators.The demographic data were collected, including age and sex, and anthropometric data were measured when undressed, including height, weight, and waist circumference (WC).Body mass index (BMI) (kg/m 2 ) is calculated by dividing the weight (kilograms) by the square of height (meters).Blood pressure monitoring is administered in a resting state.The participant takes a sitting position, and elbows are naturally placed approximately parallel to the heart to detect the systolic blood pressure (SBP) and diastolic blood pressure (DBP) of the right brachial artery.

Ultrasonography and Definitions.
Professional sonographers performed the abdominal ultrasound for the liver, gallbladder, bile duct, pancreas, spleen, and kidney.The ultrasonographic images of GSD are as follows: one or more strong echoes with acoustic shadows in the gallbladder cavity, extrahepatic bile duct, or intrahepatic bile duct, which can move with the change of body position [20].Cholecystectomy was defined as a history of gallbladder removal operation and/or no gallbladder visible at ultrasonography due to gallstones [21].GSD was diagnosed as the presence of gallstones and/or cholecystectomy [22].
2.5.Systematic Review and Meta-Analysis.The meta-analysis has been registered at PROSPERO (CRD42020179076).We systematically searched Embase and Pubmed databases to identify relevant English publications until July 2021.In addition, we searched the China National Knowledge Infrastructure (CNKI) database and Wanfang Data Knowledge Service Platform to obtain qualified Chinese Studies.The keywords include "gallstones" OR "gallstone disease" OR "cholelithiasis" AND "alanine aminotransferase" OR "aspartate aminotransferase" OR "alkaline phosphatase" OR "bilirubin" OR "serum protein" OR "gamma-glutamyl transferase" OR "risk factors" OR "influencing factors."Two authors independently performed a literature search, article selection, data extraction, and quality assessment.All inconsistent data were discussed and resolved by the corresponding authors.In addition, we searched the references of eligible studies to include additional studies.

2
Gastroenterology Research and Practice Inclusion criteria were as follows: (1) observational studies, including cross-sectional, case-control, or cohort studies; (2) investigating the relationship between liver functionrelated indicators and GSD risk in the general population; (3) provide relative risk (RR), odds ratio (OR), or hazard ratio (HR) with a 95% confidence interval (CI), or provide sufficient data to calculate these risk estimates; (4) provide standard mean difference (SMD) with 95% CI values between GSD patients and health control or provide sufficient data to calculate the SMD and 95% CI.When multiple studies focus on the same population, we only included the most comprehensive study with the largest sample size.
Data were extracted from a predesigned table, including the first author, publishing year, study period, exposure factors, geographical background, study design, sample size, the number of male and female participants, and the average level of liver function-related indicators in GSD and non-GSD groups.We extracted the most fully adjusted estimates when multiple adjusted models were used in the study.As recommendations, the quality of the included studies was assessed according to the Newcastle-Ottawa Scale (NOS) [23].A score of ≥ 7 was deemed high quality, while a score of < 7 was regarded as low quality.
2.6.UKB.The UKB is a large prospective cohort study established to investigate various diseases' genetic and lifestyle determinants, having the characteristics of large sample size, rich variables, and high data quality.In the UKB, a total of 500,000 volunteers between the ages of 40 and 69 were enrolled from the United Kingdom, which includes demographic characteristics such as age and gender; serological test indicators such as FPG and TC; behavior and lifestyle indicators such as smoke and drink; body measurement indicators, such as height and weight; disease information, such as diabetes and hypertension; abdominal ultrasound examination, such as GSD and nonalcoholic fatty liver disease (NAFLD) [19].The application number of this research is Project50538.2.7.Statistical Analysis.Continuous variables were described as mean ± SD, and differences between GSD and control group were compared by Student's t-test.Categorical variables were expressed in frequency (percentage), and differences between groups were compared by the chi-square test or Wilcoxon's rank-sum test.
For each center of the CMPEC, the relationship between liver function-related indicators and GSD risk was assessed by multivariable logistic regression analysis adjusting for confounding factors, including age, gender, BMI, kidney stones, fatty liver, high blood pressure, blood lipid levels (TC and TG), FBG, blood pressure levels (SBP and DBP).In addition, we conducted subgroup analyses by age (< 40, vs. 40-60, vs. > 60 years) and gender (male vs. female).For the systematic review and meta-analysis, when the included studies provide the OR and 95% CIs, meta-analysis for the pooled OR was assessed to examine the associations.For the included studies that offer the mean levels of liver function-related indicators and SD, meta-analysis for continuous variables was conducted to obtain standard mean difference (SMD).The I 2 was used for quantifying the amount of variance.The DerSimonian and Laird randomeffects model was used for the systematic review and metaanalysis when I 2 ≥ 50%.Otherwise, the fix-effect model was applied.The study design, geographic background, and NOS score also performed subgroup analyses.In addition, sensitivity analysis by sequentially omitting each study was conducted to assess the stability of the pooled results.Potential publication bias was evaluated by a significant funnel plot asymmetry and Egger's test.
For the UKB, the relationship between liver functionrelated indicators and GSD risk was assessed by multivariable Cox regression analysis, adjusting for confounding factors, including age, gender, BMI, WHR, Type I diabetes, Type 2 diabetes, nonalcoholic fatty liver, cirrhosis, hypertension, dyslipidemia, blood lipid levels (TC, triglycerides, HDL, and LDL), kidney stones, chronic kidney disease, cystatin C, apolipoprotein A, apolipoprotein B, C-reactive protein, and uric acid.
All statistical analyses were performed by using SPSS 26.0 (IBM, USA).Meta-analysis was performed by Stata 16 (Stata, College Station, TX).The statistically significant level was two-tailed and was set at p < 0 05.

Baseline Characteristics of Recruited Subjects in CMPEC.
A total of 369,931 participants in the CMPEC were included, and 173,050 of them were males (46.8%); the average age was 42.35 years.For different centers, the GSD prevalence ranged from 3.48% in Chongqing Qianjiang Central Hospital to 8.35% in the People's Hospital of Kaizhou District of Chongqing.Meta-analysis suggested that the pooled GSD prevalence was 5.68% (95% CI: 4.20%, 7.64%).The demographic and clinical characteristics are listed in Table 1.
Compared with the non-GSD group, the GSD group had higher s BMI, DBP, SBP, FBG, TC, TG, and LDL-C levels and a higher proportion of fatty liver disease, hypertension, kidney stones, and gallbladder polyps than the non-GSD group.While the HDL-C, ALB, and TP levels were relatively lower.

The Associations Between Liver Function-Related
Indicators and GSD Risk in the CMPEC.Multivariable logistic regression analysis showed the associations between AST, ALT, TBil, and GSD risk were inconsistent between different centers in the CMPEC.The results are listed in Table S2.
analysis also indicated that the associations were generally consistent when stratified by study design, NOS score, and geographic background.(Table 4) In some indicators, there was heterogeneity among the included studies, and meta-regression showed that the study design and geographic background might be the sources of heterogeneity (Table 4).Sensitivity analysis (Figure S2) shows that when a single study is omitted, some indicators results have changed, but most indicators results have no noticeable change.In addition, Egger's test showed no significant difference in publication bias (the funnel charts are shown in Figure S3).

Baseline Characteristics of Recruited Subjects in UKB.
A total of 462,903 participants in the UKB were included, and 210,741 of them were males (45.5%), and the average age was 56.76 years.The median follow-up time was 150 months, and the interquartile interval was 20 months.During the follow-up, a total of 15,071 participants (2.66%) developed GSD, and 6842 of them were males (45.4%) (Table S1).The multivariable Cox analysis in the UKB showed that ALP (HR = 1 09, 95% CI: 1.07-1.11,p < 0 001) was positively associated with GSD risk.In contrast, AST (HR = 0 95, 95% CI: 0.94-0.977,p < 0 001) was negatively related to the risk of GSD.(Table 5).

Discussion
In our study, we first conduct a cross-sectional study through CMPEC to find the associations between liver function-related indicators and GSD risk.After that, we conduct a systematic review and meta-analysis to confirm these associations.And then, we combined the abovementioned results by meta-analysis to get a relatively definitive conclusion.Ultimately, we verify these associations based on the subjects in the UKB cohort.
Results for the CMPEC and the systematic review and meta-analysis found a high level of AST, ALT, and ALP were positively associated with GSD risk, and high TP and ALB levels were negatively associated with GSD risk.While in the UKB, only the association between ALP and GSD was validated.Besides, AST was negatively associated with GSD risk in the UKB.
Our study reveals a compelling positive correlation between ALP and the risk of GSD, suggesting that higher levels of ALP may significantly contribute to an increased susceptibility to developing GSD.The mechanism remains unclear, possibly because when hepatic cells are injured, ALP, a biomarker of hepatobiliary disease and bone resorption, is commonly present in liver and bile duct cells.This may increase the pressure in the bile duct and reduce bile excretion, thereby increasing the risk of GSD [46].Therefore, the level of ALP can be used as an effective indicator for clinical diagnosis and prediction of gallstone onset.
ALB, as a crucial constituent of TP, is synthesized by liver parenchymal cells at a rate contingent upon colloid osmotic pressure and dietary protein intake, thereby capable of reflecting the liver functional reserve [47].Our research indicates that high levels of TP and ALB are negatively correlated with the risk of GSD.However, this association was not validated by the UKB, and more studies are warranted to further confirm these results.
AST and ALT are markers that reflect hepatocyte damage and enter the bloodstream during hepatocyte necrosis or increased hepatocyte membrane permeability, with increased enzyme activity in the blood [48,49].The results of the multicenter study and the systematic review and meta-analysis showed that AST and ALT were positively associated with the risk of GSD.While in the UKB, AST was negatively associated with GSD risk, and ALT was not significantly associated with GSD.Cross-sectional studies in Chinese populations concluded that AST was not significantly associated with GSD [50].One study [17] also showed no association between ALT and the risk of GSD.That may be because ALT is more sensitive or specific to the liver, resulting in unstable analysis results of ALT.The elevation of AST may also be considered secondary to nonhepatic causes, but the exact mechanism of AST remains unclear.In our study, the meta-analysis results of the association between AST and GSD were inconsistent with those obtained by UKB.In the future, we need more multicenter-cohort studies to further verify the results.

4.1.
Limitations.There are some limitations to our study.Firstly, only cross-sectional data of the CMPEC is available for this part, and the findings may be influenced by potential confounders and reverse causality.Further prospective studies are subsequently needed to verify the association between liver function-related indicators and GSD risk.Secondly, most of the studies included in the meta-analysis were from China, which may be subject to some bias, due to the unavailability of the original data of the included studies, most of the studies could not be meta-analyzed for OR values.To minimize this bias, we performed a meta-analysis of SMD values for all studies (Table 3).Finally, in this study, we only studied gallstones and did not include choledocholithiasis or other gallbladder diseases, and there is a need for conducting further studies on other gallbladder diseases in the future.

Conclusion
In conclusion, our findings show that ALP, low TP, and low ALB are risk factors for GSD, which can provide a basis for 11 Gastroenterology Research and Practice identifying the high-risk groups of GSD and exploring the pathogenesis of GSD.When the indexes of the liver function of patients are abnormal, clinical attention should be paid to them.Besides taking effective treatment and preventive measures, health publicity and education should be strengthened to improve self-care to reduce the incidence of GSD.In addition, monitoring serological indexes related to liver function can provide a basis for identifying high-risk groups, early predicting, and improving prognosis.

2. 1 .
Study Population.CMPEC consisted of nine hospitals.The present study included the participants who underwent health check-ups in Chongqing Kaizhou District People's Hospital, First Affiliated Hospital of Chongqing Medical University, Beijing Xiaotangshan Hospital, Tianjin Medical University Cancer Institute and Hospital, and Chongqing Qianjiang Central Hospital in China between January 2015 and May 2020.Participants aged ≥ 18 underwent abdominal ultrasonography, had laboratory results of liver functionrelated indicators, and signed informed consent were included.The Ethics Committee of West China Fourth Hospital and West China School of Public Health, Sichuan University approved the study.(Gwll2021055).

Figure 1 :
Figure1: Flow chart for systematic review and meta-analysis.

Table 1 :
Baseline characteristics of recruited subjects according to gallstone disease status.

Table 2 :
Pooled analysis of the association between liver function-related indicators and gallstone disease risk in the multicenter study.

Table 5 :
The association between liver function-related indicators and GSD risk in the UKB.