Abstract
Objective: Only few determinants of gallstone formation have been identified in cohort studies. The aim was to identify further determinants for gallstones in a Danish cohort and to perform a meta-analysis of results from existing cohorts.
Material and methods: Data from a cohort study was used. Gallstone incidence was assessed through repeated ultrasound examinations. Body mass index (BMI), blood pressure, self-rated health, lifestyle variables, blood lipids, and use of female sex hormones were measured at the baseline examination. Statistical analyses included logistic regression. Based on a prospective protocol, a systematic review of the literature was performed identifying all articles dealing with determinants of incident gallstones. Meta-analyses of comparable determinants were performed through fixed effect models.
Results: Participants with no gallstones at baseline and with at least one re-examination were followed-up completely (mean 11.6 years, N = 2848). The overall cumulative incidence of gallstones was 0.60% per year. Independent positive determinants for incident gallstones were age, female sex, non-high density lipoprotein (non-HDL) cholesterol, and gallbladder polyps. In addition, BMI was positively associated in men. The systematic review additionally identified associations for comorbidities, parity, and dietary factors. Meta-analysis confirmed the significant associations for incident gallstones and age, female sex, BMI, and non-HDL cholesterol. No significant associations were found for blood pressure, smoking, alcohol consumption, HDL cholesterol, or triglycerides in meta-analyses.
Conclusions: Age, female sex, BMI, non-HDL cholesterol, and polyps are independent determinants for gallstone formation. Incident gallstones and the metabolic syndrome share common risk factors. More studies are needed for further exploration.
Acknowledgements
We would like to thank Sarah Louise Klingenberg for her work with the search strategy, Anja Lykke Madsen for data management, Rikke Kart Jacobsen for statistical supervision, and Karin Mønsted Shabanzadeh for linguistic edit.
Disclosure statement
The authors declare no conflicts of interest. This work was supported by The Faculty of Health and Medical Sciences, University of Copenhagen, Denmark.