Objectives Gallstone disease (GSD) can be caused by various health and clinical factors such as obesity, dyslipidaemia and an unhealthy diet, all of which are associated with higher high-sensitivity C reactive protein (hs-CRP) concentrations. Whether hs-CRP represents an independent risk factor for GSD is still unclear. We prospectively investigated hs-CRP in relation to the occurrence of GSD based on the Kailuan study.
Study design Prospective cohort study.
Setting The Kailuan cohort study was conducted in Tangshan City in northern China.
Participants 95 319 participants who were free from GSD were recruited in this study. Epidemiological data, anthropometric parameters and biochemical data of participants were collected.
Primary and secondary outcome measures Cox proportional hazards regression models were used to evaluate the association between hs-CRP concentrations and the risk of GSD after adjustments for potential confounders.
Results During the mean 7.58 years of follow-up among 95 319 participants, 4205 participants were identified as newly diagnosed with GSD or having undergone cholecystectomy for cholelithiasis. Compared with the hs-CRP<1 mg/L group, elevated hs-CRP concentrations were significantly associated with higher risk of GSD with the corresponding HR of 1.11 (95% CI 1.03 to 1.19), 1.12 (95% CI 1.04 to 1.22) in the 1≤hs-CRP≤3 mg/L and hs-CRP>3 mg/L group, respectively. The multivariate model which included hs-CRP not only had a better line of fitness but also had better predictive values to help identify new cases of GSD during follow-up.
Conclusion Elevated hs-CRP concentration is an independent risk factor for new-onset GSD among the Chinese population.
Trial registration number ChiCTR-TNC-11001489.
- motility disorders
- public health
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
Ultrasound examination was used in this study and was able to represent the true incidence of gallstone disease (GSD) in a defined population.
This study is a large-scale population-based study, with a high number of new-onset GSD cases.
Distinguishment between the two different types of stones (cholesterol and pigment) could not be carried out via ultrasonographic examination.
GSD incidence was higher in women, but no data concerning the number of pregnancies or the frequency of oral contraceptives were used in our study.
Gallstone disease (GSD) is the most common disease of the biliary tract and the leading cause of gastrointestinal-related inpatient admissions.1 The prevalence and incidence of GSD appear to be higher in developed countries than in developing countries like China.2–4 The third National Health and Nutrition Examination Survey estimated 20 million adults (6.3 million men and 14.2 million women) in the USA are afflicted with GSD, with the cost of treatment estimated at 6.2 billion annually.5 6 GSD increases the incidence of all-cause mortality as well as cancer and cardiovascular-related mortality.7 8 The pain caused by GSD in the right upper quadrant can seriously impact quality of life. Previous cohort studies have demonstrated that gallstones found in the general population can cause symptomatic disease in 18%–24% of patients. In addition, 3%–8% of patients suffer from complications like acute cholecystitis, gallstone ileus, pancreatitis, gallbladder empyema or perforation, causing GSD to be a major public health problem.9–12
Several risk factors for GSD are well established, including ageing, female gender, ethnicity/genetics, pregnancy, family history of GSD and sedentary lifestyle.13–16 Previous studies have illustrated that inflammation is closely related with cardiovascular disease, atherosclerosis and hypertension. C reactive protein (CRP) is generally measured in inflammation as a non-specific acute-phase protein. CRP and high-sensitivity C reactive protein (hs-CRP) are essentially the same, with the main difference between them being analytical sensitivities and assay ranges.17–19 Besides the aforementioned chronic diseases, elevated hs-CRP has been proven to be linked to various health metrics and clinical factors such as obesity, dyslipidaemia, unhealthy diet, diabetes and lack of physical activity,20–24 all of which may predispose the individual to GSD. Whether elevated hs-CRP concentration represents an independent risk factor for GSD after traditional risk factors are taken into account is still unclear. We therefore conducted this study to prospectively explore potential associations of hs-CRP with new-onset GSD based on the Kailuan cohort study.
Materials and methods
Patient and public involvement
Patients and the public were not involved in the development, design or analysis of this study.
The data were obtained from health examinations of employees of the Kailuan Company in Tangshan city, which is situated 150 km southeast of Beijing. The Kailuan Company covers a wide range of industries and fields, including coal products, machine building, electric power, chemical production, transportation, construction and healthcare. The Kailuan study, a prospective population-based study, was designed to investigate risk factors for chronic diseases. From July 2006 to October 2007, all 155 418 employees (including retired employees) of the Kailuan Company were invited to participate. A total of 101 510 participants (65.3%) agreed, and written informed consent was obtained. These participants underwent complimentary routine medical examinations every 2 years. All examination reports were reported to participants after 1 week.
A total of 101 510 working and retired employees from the Kailuan Corporation underwent the baseline examination from July 2006 to October 2007 at Kailuan General Hospital and its 10 affiliated hospitals. Information including physical examination, type-B ultrasound examination, blood sample, urine sample and biochemical tests was collected. Questionnaires were done via face-to-face interviews by the medical staff and trained research nurses. Participants were given repeated questionnaires and underwent medical examinations biennially. In the current study, we excluded 4974 participants with a history of GSD (4737 stones cases, 237 cholecystectomy cases), 757 with data missing from type-B ultrasound examination, 954 participants without measurement of hs-CRP and a final 860 participants without data of Body Mass Index (BMI, in kg/m2), total cholesterol (TC, in mmol/L), triglyceride (TG, in mmol/L), fasting blood glucose (FBG, in mmol/L), systolic blood pressure (SBP, in mm Hg) and diastolic blood pressure (DBP, in mm Hg), alcohol and tobacco consumption, and exercise frequency. Due to environmental metabolic disorders in patients with malignant tumours, we also excluded 367 participants with a history of cancer. Subsequent biennial follow-up examinations were conducted until the endpoint of this study (31 December 2015). A total of 6272 participants failed to complete the series of assessments. A total of 87 326 participants who participated in at least one follow-up examination were ultimately recruited in this study. Participant screening details are shown in figure 1.
Epidemiological survey and anthropometric parameters
All patients underwent an interview with a standardised questionnaire which included questions on smoking status, level of physical activity and drinking intake. Anthropomorphic parameters including height, weight and waist circumference (WC) were measured.25 26 The BMI was calculated as the ratio of body weight (kilogram) divided by the square of body height (metre). A minimum of a 5 min rest, three readings of SBP and DBP were taken by using a mercury sphygmomanometer with a cuff of appropriate size. The average of three readings was used for the analysis. Hypertension was defined as the presence of any of the following: SBP≥140 mm Hg and/or DBP≥90 mm Hg, a history of hypertension or the usage of antihypertensive treatment. Smoking status was defined as having at least one cigarette every day in the recent year. Drinking status was defined as having taken 100 mL/day (alcohol contents>50%) of alcohol on average for more than 1 year, and physical exercise as exercising ≥3 times/week, ≥30 min/time.27 28
Determination of plasma hs-CRP levels and other biochemical indicators
Overnight fasting (at least 8 hours) blood samples were collected from the cubital vein at 07:00–21:00 on the physical examination day and transfused into vacuum tubes containing EDTA. TG and TC were both measured using the enzymatic method. FBG was measured using the hexokinase/glucose-6-phosphate dehydrogenase method. Plasma hs-CRP was measured using a commercial, high-sensitivity particle-enhanced immunonephelometry assay (Cias Latex CRP-H, Kanto Chemical Co., Tokyo, Japan) with a detection limit of 0.1 mg/L. In accordance with the guidelines from the Centers for Disease Control and Prevention and the American Heart Association, participants were categorised into three groups based on their hs-CRP concentration: hs-CRP<1 mg/L, 1 mg/L≤hs-CRP≤3 mg/L and hs-CRP>3 mg/L.29 All the plasma samples were analysed using an autoanalyzer (Hitachi 747; Hitachi, Tokyo, Japan) at Kailuan General Hospital central laboratory. Diabetes was defined according to the criteria of the American Diabetes Association: FBG 7.0 mmol/L and/or a self-reported history of diagnosis, and/or previous or current hypoglycaemic therapy.
Determination of GSD
The presence of GSD at the baseline examination and new-onset GSD cases during all follow-up examinations were assessed by ultrasound examination. PHILIPS HD-15 colour Doppler ultrasound scanner with a probe frequency of 3.5 MHz was used to diagnose GSD cases. After at least 8 hours of fasting, ultrasound examination was performed in the abdominal region (examination of liver, gallbladder, pancreas and spleen in supine position). Cases of GSD excluded gallbladder polyps and included only presence of gallbladder stones and history of cholecystectomy, which were based on the presence of ‘movable hyperechoic foci with acoustic shadow’ and whether gallbladder removal due to cholelithiasis had occurred.
Data input was carried out by trained nurses of each participating hospital and transferred to the Oracle10g database located in Kailuan Hospital. Continuous and normally distributed variables were presented as means±SD and compared using one-way analysis of variance. Data in the skewed distribution were described by median±IQR and analysed using non-parametric tests. Categorical variables were presented as n (%) and compared using χ2 test. Logarithmic transformation was used to incorporate hs-CRP for analyses with continuous variables to diminish the effect of extreme observations. Person-years of follow-up were calculated as the ending point of the first examination until the diagnosis of new-onset GSD or censoring, or end of follow-ups, or GSD unrelated death, whichever occurred first. We used Cox proportional hazard model to calculate HRs and 95% CIs for new-onset GSD by using log(hs-CRP) in the regression model as a continuous variable. Several multivariable Cox proportional hazard models were adjusted for the effects of confounding covariates. Similar analytical methods were used to test the effects of the three hs-CRP groups on the risk of GSD. To identify the association between hs-CRP and new-onset GSD, subgroup analyses were carried out by stratifying participants according to sex, age (in years) (youth: ≤45, middle age: 45–59 and old age: ≥60) or BMI (in kg/m2) (normal:＜24, overweight: 24~28, obese: ≥28). Akaike information criterion (AIC) was used as a measure of the goodness of fit of the multivariate models. The discriminatory capability of the two multivariate models was evaluated using c statistics. In the sensitivity analysis, 3515 participants with hs-CRP>10 mg/L were further excluded to eliminate the effect of acute inflammatory response. Reported p values were two-sided, and p<0.05 was recorded as significantly different. All statistical analyses were performed using SAS V.9.4 statistical software.
The mean age was 50.78±12.01 years old with 75 141 (78.83%) men and 20 178 (21.17%) women. The baseline characteristics for participants stratified by hs-CRP levels are shown in table 1. Participants with higher hs-CRP concentrations were older, had larger WC, and higher SBP, DBP, TC, FBG and white blood cell count. Elevated hs-CRP concentration groups were associated with higher prevalence of hypertension and diabetes, and lower prevalence of drinking status, smoking and exercising. It is interesting to note that of the three groups, the concentrations of TG, TC, alanine aminotransferase (ALT) and the prevalence of drinking status, smoking status and physical activity were the highest in participants with hs-CRP of 1–3 mg/L.
The follow-up period ended on 31 December 2015, with a median of 7.59 years for 87 326 participants. A total of 4205 participants were newly diagnosed with GSD or had underwent cholecystectomy for cholelithiasis. It is worth noting that 3460 participants died before the end of follow-up or the occurrence of GSD. These participants were also included as they were involved in at least one follow-up examination. Table 2 shows the crude and adjusted HRs of new-onset GSD for per unit of log(hs-CRP). The HR for the association between log(hs-CRP) and GSD was 1.14 (95% CI 1.12 to 1.17) in the univariate model, 1.06 (95% CI 1.04 to 1.08) after adjusting for age and sex, and 1.03 (95% CI 1.01 to 1.05) after adjustment for sex, age, TC, TG, BMI, TC, TG, ALT, smoking, drinking status, diabetes, hypertension and physical activity.
Table 3 displays the incidence of GSD and the association of the three prespecified hs-CRP groups with GSD. The crude incidence of GSD per 1000 person-years was 7.10 (7.55 in women, 6.96 in men). The age-standardised and sex-standardised overall incidence of GSD in this study was 6.99 per 1000 person-years. Our study indicated a clear trend based on hs-CRP concentrations, with a positive correlation between hs-CRP and age-standardised and sex-standardised incidence of GSD. The incidence of GSD was 5.96 per 1000 person-years, 7.99 per 1000 person-years and 9.13 per 1000 person-years in the hs-CRP<1 mg/L, 1 mg/L≤hs-CRP≤3 mg/L and hs-CRP>3 mg/L groups, respectively. In crude models, elevated hs-CRP concentrations were significantly associated with higher risk of GSD. After adjusting for other aforementioned confounding factors, risk of GSD increased with hs-CRP levels, with the corresponding HR of 1.11 (95% CI 1.03 to 1.19) and 1.12 (95% CI 1.04 to 1.22) in the 1≤hs-CRP≤3 mg/L and hs-CRP>3 mg/L groups, respectively.
Table 4 shows the effect of hs-CRP on new-onset GSD after stratifying the participants by sex, age or BMI. In multivariate analysis, the effect of hs-CRP on new-onset GSD was not modified by sex. There was a significant positive tendency of elevated risk for GSD as hs-CRP increased in both men and women after adjusting for aforementioned covariates, and the p value for interaction was 0.143. Similar results were also observed when participants were stratified by BMI, and the p value for interaction was 0.157. However, when stratified by age, HR values increased only in the youth and middle-aged groups and not in the elderly group, and the p value for interaction was <0.001.
The AIC and c-statistic for GSD outcomes according to different prediction models were also calculated in this study. The traditional multivariate model contained several risk factors including age, sex, BMI, TC, TG, ALT, smoking status, drinking status, diabetes mellitus, hypertension and physical activity. The improvement of fit and predictive value were analysed after adding hs-CRP to the multivariate model. AIC for the traditional multivariate model and multivariate+hs-CRP model was 88 160.76 and 88 152.88 respectively. The c-statistics for the traditional multivariate model and multivariate+hs-CRP model were 0.681 and 0.693, respectively, for GSD, showing a small improvement.
To avoid the effect of acute inflammation, we further excluded participants with hs-CRP greater than 10 mg/L in the sensitivity analyses. Similar results were also observed where risk of GSD increased with the level of hs-CRP (see online supplemental table S1).
To our knowledge, this prospective study is the first to analyse the relationship between circulating levels of hs-CRP and the risks of new-onset GSD. The results indicated that elevated hs-CRP concentration was significantly associated with increased risk of GSD in both continuous variable analyses and categorical analyses among 95 319 participants.
The overall age-standardised and sex-standardised incidence of GSD was higher compared with other Chinese population-based studies, but still lower than those based on Western populations.30 31 The incidence of GSD varies greatly among different ethnic groups as well as geographical regions. This disparity might be due to the different occurrence of two distinct types of gallstones. Cholesterol stones are common among Western countries, while pigment stones are common in Asia.32 33 We speculate that continued increase in the Westernisation of diets and decrease in physical exercise may cause the prevalence and incidence of GSD in the Chinese population to reach the level of its Western counterpart within a short period of time.
Women demonstrated higher incidence compared with men, which was in line with previous research that demonstrated women were almost twice as likely to develop gallstones.34 Female predominance in the incidence of GSD could be due to reduced hepatic bile secretion, as female sex hormones increase cholesterol secretion and diminish bile salt secretion.35
In this study, we found that an elevated hs-CRP concentration was associated with a higher risk of GSD in univariate analysis. After adjusting for other potential confounding variables, the association between hs-CRP and GSD was attenuated but remained significant. Participants with hs-CRP>3 mg/L had 12% increased risk of GSD versus the lowest hs-CRP level. Similar results were also obtained when participants were stratified by sex or age. However, our study failed to find a positive association of hs-CRP with GSD in elderly people. GSD is rare in neonates and young people, and long-term exposure to risk factors could explain the increased incidence of GSD in the elderly.36 In previous studies, older participants were found to have higher hs-CRP concentrations,37 and we thus expected to find higher risk of incidence. We currently have no clear explanation for such discrepancy. Further studies should be conducted to investigate whether age eliminates the risk of CRP for GSD among the elderly. Shabanzadeh et al38 demonstrated that CRP was significantly associated with GSD (OR 1.03, 95% CI 1.01 to 1.05) among 2650 Danish participants which is partially supportive of our findings.
Traditional risk factors such as female gender, ageing, obesity and diabetes mellitus have been demonstrated in previous studies concerning the risk of GSD globally.39 New risk factors should be evaluated in prospective cohort studies to allow for further accuracy in estimation of the actual risk, leading to the most appropriate care and treatment. Our results revealed that the multivariate+hs-CRP model had better fitness and also a better predictive value to help identify new cases of GSD during follow-up. Taking hs-CRP concentration into consideration may be a better method to identify individuals’ risk of GSD.
The mechanisms that cause elevation of hs-CRP with increased risk of GSD remain uncertain. Obesity is considered to be closely related to the formation of stones, and previous research suggests that adipose tissues stimulate the secretion of interleukin-6, which is involved in the production of CRP in hepatocytes.37 Berk et al found that chronic inflammation of any aetiology might lead to calcium being deposited in the gallbladder wall, which may explain the pathogenesis between CRP and GSD.40 In animal experiments, cholesterol crystals were closely associated with inflammation of the gallbladder wall,41 as well as increased gallbladder wall thickness.42 Elevated gallbladder wall thickness may decrease motility,42 suggesting that gallbladder inflammation might cause gallbladder dysmotility. Additional investigations should explore the pathophysiological mechanisms among CRP, cholecystitis and GSD, as well as whether elevated CRP is an indicator of GSD, rather than a sign of acute or chronic cholecystitis.
Ultrasonography has been regarded as the gold standard method for the diagnosis of GSD. The positive predictive value has been reported as 99%–100% and the negative predictive value as 90%–96%.43 The availability of ultrasound has facilitated the conduction of epidemiological surveys in Western countries.44 Ultrasound examination was also able to represent the true incidence in a defined population as it eliminates the bias of earlier research: necropsy implies death and clinical diagnosis depends on symptoms (80% of the stones are clinically silent45).
The current study is a large-scale population-based study, with a high number of new-onset GSD cases. There were several strengths of this study, including sonography evaluation of the gallbladder, a large sample size and high participation rates during a nearly 8-year follow-up period, and a broad assessment of potential confounders. However, potential limitations should also be noticed. First, we could not distinguish between the two different types of stones (cholesterol and pigment) and the condition of stones by ultrasonographic examination. Second, the incidence was higher in women, but no data concerning the number of pregnancies and the frequency of oral contraceptives were used in our study. Third, there was an imbalance in sex distribution due to the industrial nature of the Kailuan Community. However, the influence of imbalance in sex distribution on the results would be minimal as we carried out independent statistical studies on both genders. Fourth, because of the industrial nature of Kailuan community, a large proportion of participants are labour workers. The results may not be a true representative of the general Chinese population.
The main implications in clinical practice would be an increased awareness of hs-CRP and its correlation to the risk of gallbladder disease. We would like to point out that this is, hopefully, a stepping stone to other insightful research regarding chronic inflammation and gallbladder disease.
Our study confirmed that elevated hs-CRP concentration was an independent risk factor for new-onset GSD among the Chinese population.
We thank the staff and participants of the Kailuan study for their important contributions.
TL and STS are joint first authors.
JQ and SL contributed equally.
Contributors TL and STS executed the study and drafted the manuscript. TL, STS and NY participated in the study design and performed the statistical analyses. GX, WL and ND contributed to the discussion. SL and JQ reviewed the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Patient consent for publication Obtained.
Ethics approval This study was approved by the ethics committee of Kailuan General Hospital and was in compliance with the Declaration of Helsinki. Informed consent forms were signed by the participants or their legal representatives.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement No data are available.