Article Text
PDF
Abstract
Objective To investigate factors associated with the prevalence and incidence of gallstone disease (GSD) in women and men of the MAUCO population-based prospective cohort.
Design 8948 MAUCO participants (aged 38–74 years) underwent abdominal ultrasound at baseline (2015–2019); 4385 received follow-up ultrasound at years 2 or 4. Factors associated with prevalent GSD were assessed using Poisson multiple regression and with incident GSD using Cox regression models.
Results GSD prevalence was 40.4% in women (13.1% gallstones, 27.3% cholecystectomies) and 17.1% in men (8.9% gallstones, 8.2% cholecystectomies). In men, GSD prevalence rate ratio (PRR) by age in >64 years was 3.85 (95% CI 3.00 to 4.94), doubling that of women’s PRR 1.78 (95% CI 1.57 to 2.01). In women, waist circumference and diabetes were stronger GSD factors; a higher number of children and worse metabolic and socioeconomic conditions were also highlighted. GSD men had higher cardiovascular disease and a family history of GSD and gallbladder cancer. 198 GSD cases developed during follow-up, with incidence increasing by 2% (95% CI 1.005% to 1.03%) per each centimetre above the ideal waist circumference, statistically significant only in women. In men, age was the strongest factor for incidence, followed by a family history of GSD and low high-density lipoprotein increased incidence risk.
Conclusions GSD burden was high in this population; a third of women had their gallbladder removed, which may pose them at risk of other health problems. Abdominal obesity was the only preventable GSD risk factor, highlighting the need for effective public health policies promoting obesity reduction.
- GALLSTONE DISEASE
- GALLSTONES
- CHOLECYSTECTOMY
- OBESITY
Data availability statement
Data are available upon reasonable request.
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
WHAT IS ALREADY KNOWN ON THIS TOPIC
Chile has among the world’s highest prevalence of gallstone disease (GSD) and mortality from gallbladder cancer. However, no prospective study has investigated its natural history in the country.
WHAT THIS STUDY ADDS
We found strong evidence that having older age, higher waist circumference, moderate-to-severe fatty liver and family history of GSD were associated with having GSD.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This study confirms that the Chilean population has one of the highest GSD burdens worldwide, with a higher prevalence in women (40.4%) than in men (17.1%). To reduce GSD prevalence, public health programmes should focus on reducing obesity, the only modifiable risk factor identified in this study.
Introduction
Gallstone disease (GSD), including gallstones and cholecystectomy, affects 10–15% of adults in developed countries,1 and its prevalence is increasing among younger individuals in line with the dramatic increase in obesity.2–4 Gallstones are the leading risk factor for gallbladder cancer.5–7 Chile has among the world’s highest prevalence of GSD and mortality from gallbladder cancer.8–10 Obesity, unhealthy diet, diabetes, non-alcoholic fatty liver disease and genetic variables are the main risk factors for GSD.11 12 A recent meta-analysis of 436 636 elective cholecystectomies worldwide found an average prevalence of incidental gallbladder cancer of 0.6% (95% CI 0.5% to 0.8%).13 Similarly, a study in Turkey reported incidental gallbladder cancer in 1%, ranging from 0.3% among individuals <60 years of age to 2.6% in those ≥60 years.14 In Valdivia and Temuco, the high-risk areas for gallbladder cancer in Chile, researchers found a 4% incidental gallbladder cancer between 1990 and 2010 (personal communication Dr Enriqueta Bertrand and JCA), figures much higher than the 0.23% reported by Bragheto in 1999 in Santiago, the capital of the country and a lower risk area.15 Considering that studies in the field are limited and that the worldwide prevalence of GSD is evolving, this manuscript fills gaps in the literature and includes information relevant to other countries facing this new reality. The main aim of this study was to describe the prevalence and risk factors associated with GSD by sex in a population-based cohort of chronic diseases (the Maule Cohort (MAUCO)).16 17 Moreover, to increase our understanding of the long-term effects of GSD, we are also investigating the association between risk factors and GSD incidence in the same cohort by sex. This study is the largest population-based cohort of ultrasound-detected GSD in Latin America.
Methods
Study design, setting and selection of participants
The MAUCO target population includes residents of Molina County in Central Chile, which is representative of small counties of low socioeconomic levels. This population is covered by public health insurance and has a primary health hospital, two health centres and two rural health posts. Reference hospitals are between 20 and 56 km from Molina.
Through a household census, we invited all adults aged 38–74 years to enrol, excluding those unable to consent autonomously or who were terminally ill16 17; 72.3% accepted. Further methodological details are described elsewhere.16 17 Between 2015 and 2019, individuals who accepted participation signed consent forms and completed a health and lifestyle survey (personal and family medical history, medication use, cardiovascular and digestive symptoms and neurocognitive state). They underwent anthropometric measurements, bioimpedance analysis and abdominal ultrasound at the study clinic as well as provided blood and saliva samples.17 All participants received follow-up surveys 2 years later, and a subgroup of the cohort—those with abnormal baseline ultrasound and approximately 1:1 age-sex matched controls with normal baseline ultrasound—were invited for a new ultrasound exam (cohort control). The follow-up rate for ultrasound at 2 years was 96%.17 Additionally, any participant who visited the MAUCO clinic was offered a follow-up ultrasound (opportunistic controls). Participants with abnormal health results at baseline or follow-up were referred to Molina Hospital. 8948 participants were finally included in the main baseline analyses (online supplemental figure 1). The distribution of missing variables at baseline by sex is available in online supplemental table 1.
Supplemental material
Gallstone disease ascertainment
A medical technician (FH), trained and supervised by a radiologist (FC) at Pontificia Universidad Católica de Chile, performed and recorded all abdominal ultrasounds following the Rumack 2011 Guidelines.18 Images were stored and reviewed by the radiologist as needed. GSD, the outcome variable, included cholecystectomy, gallstones or biliary sludge. Other abnormal gallbladder results included cholesterolosis, polyps, wall thickening, wall calcifications, scleroatrophic gallbladder, image suggestive of neoplasm, adenomyomatosis and adenomyosis (online supplemental table 2). Liver steatosis was classified as absent, mild, moderate or severe.19
Follow-up ultrasound examination
Among the 8609 MAUCO eligible for follow-up participants, 2353 participants with GSD, 1811 cohort controls and 597 MAUCO cohort opportunistic controls were invited for a follow-up ultrasound (online supplemental figure 1). The equipment and personnel were the same as at baseline.
Predictors of gallstone disease
Socio-demographic, lifestyle, familial and health-related factors were evaluated as prevalent and incident GSD risk factors. Socio-demographic factors included age at recruitment (38–44, 45–54, 55–64, 65–74 years), sex (women/men), self-reported schooling (<8, 8–12, >13 years), health insurance (levels A/B, C/D or private; where A corresponds to the lowest income level), and ancestry (self-identified Amerindian or Chilean Hispanic). Health factors included diet (Mediterranean Diet Score, defining healthy score ≥9),20 food consumption patterns (eg, ≥1 fruit per day), alcohol intake (risky ≥20 g per week in women and ≥30 g in men), self-reported current or ever smoker (≥100-lifetime cigarettes), physical activity (low: <3, 30 min sessions per week), grip strength (low <27 kg in men and <16 kg in women), self-reported walking pace (slow, normal or brisk pace) and the number of teeth (assessed by a health technician and classified as ≥20 or <20). Women’s hormonal factors included the number of children (count or dichotomised ≥3), hormonal contraception (ever) and hormonal replacement therapy (ever). Gallstones or gallbladder cancer in parents, children or full and half-siblings were also evaluated. Obesity was considered as a waist circumference ≥88 cm in women and ≥102 cm in men or body mass index (BMI)≥30 (weight (kg)/height (m2)). Abnormal lipids were defined as low-density lipoprotein (LDL) cholesterol ≥160 mg/dL, high-density lipoprotein (HDL) cholesterol <50 mg/dL in women and <40 mg/dL in men and triglycerides ≥200 mg/dL. Altered bilirubin and liver enzymes were defined as total bilirubin >1.2 mg/dL, direct bilirubin >0.3 mg/dL, aspartate aminotransferase ≥48 IU/L, alanine aminotransferase ≥55 IU/L and alkaline phosphate ≥129 IU/L. Chronic diseases considered as risk factors for GSD included diabetes (self-reported, glycaemia ≥126 mg/dL or hypoglycaemic drug use), metabolic syndrome (≥3: abdominal obesity, high triglycerides, low HDL, high blood pressure and high fasting glucose), high blood pressure (systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥80 mm Hg or use of antihypertensive medication) and self-reported cardiovascular disease (myocardial infarction, heart failure, stroke or arrhythmia). Digestive symptoms assessed were biliary colic (acute pain in epigastrium lasting >30 min without diarrhoea) and dyspepsia (Rome criteria III) in the past 12 months.
Statistical analyses
There was a very low percentage of missing data (online supplemental table 1), therefore, values were not imputed for these analyses.21–23
Prevalent gallstone disease
Factors associated with baseline GSD were tested with χ2 tests or t-tests. GSD age prevalence curves by sex were tested with the Kolmogorov-Smirnov test. We tested the models with the Akaike information criteria for risk factors and confounder selection. We ran multiple logistic regression models separately by sex to obtain the prevalence of GSD by each baseline variable. Prevalence rate ratios (PRR) and 95% CI were calculated with age-adjusted robust Poisson multiple regressions.24–26
Incidental GSD and changes in gallbladder status
Participants at risk were those without gallstones or cholecystectomy at baseline (n=6256) who had a follow-up ultrasound by September 2022 (n=2284, 36.5%); gallstones or cholecystectomy detected in follow-up were considered incident GSD (online supplemental figure 1). We calculated the time from enrolment to the follow-up ultrasound to obtain person-time at risk between ultrasounds, estimating the incidence rate of GSD per 1000 person-years. We used Hazard Ratios (HRs) from age-sex adjusted multiple Cox regression to select variables predictive of GSD incidence. To investigate the risk factors associated with GSD incidence, we first produced directed acyclic graphs based on current knowledge about the disease (online supplemental figure 2).27 28 Then, the following models were tested: (1) the overall and men model: adjusted by sex, schooling, family history of GBC, HDL cholesterol, diet, age and age x sex interaction; (2) the women model: as per the overall model but additionally adjusted for number of children. The proportional hazard assumption was checked using Schoenfeld residuals (the proportional hazard assumptions were all non-significant with a global p-value of 0.28). Finally, we assessed competing risk accounting for all-cause mortality using the Cox proportional hazard model for GSD incidence. The Fine-Gray subdistribution risk model was used to estimate the specific influence of obesity on the onset of GSD, adjusted for the aforementioned models and incorporating death as a competing event. The cumulative incidence curves for GSD and mortality, stratified by sex, were generated using the ‘cmprsk’ package in R. R V.4.5.1 and Stata V.15 (StataCorp LP) statistical software were used for the analyses.
Results
Prevalent gallstone disease
Of the 8970 (69.6%) Molina residents who participated, 8948 (99.8%) had a valid ultrasound and were included in our prevalence analyses; 2692 (30%) had GSD and 6256 (69.9%) had normal gallbladder or other anomalies (table 1 and online supplemental figure 1). GSD was twice as frequent in women (40.4%) as in men (17.1%) (table 1). The proportion of GSD cases who had already received a cholecystectomy was 67.6% in women and 48% in men. Women tended to have larger stones, while men had a higher prevalence of other anomalies (2.3% vs 1.6% in women) (table 1 and online supplemental table 2). On the other hand, of the 1193 people who reported biliary colic at enrolment, 17.9% were gallstone carriers, 21.5% had a cholecystectomy and 60.5% were participants without GSD (online supplemental table 2).
Gallbladder ultrasound findings at baseline. MAUCO 2015–2019
Baseline characteristics of the included population by GSD status and sex are available in online supplemental table 3. Overall, GSD was associated with higher markers of obesity, metabolic syndrome and fatty liver. Men with GSD had higher cardiovascular disease and family history of gallbladder disease, while women with GSD had more children than women without GSD. Women were more likely to have digestive symptoms independent from their GSD status, while men with GSD were more symptomatic than those without GSD (online supplemental table 3). Other characteristics of cases, non-GSD ultrasound cohort controls and matched and opportunistic controls, can be found in online supplemental table 4.
Factors associated with prevalent GSD at baseline by sex are presented in figure 1 and online supplemental table 5. In multivariate models, age was the strongest factor associated with GSD in both sexes (figure 1). Compared with those aged 38−44, men aged ≥65 had a 3.85-times higher GSD PRR, while women ≥65 years had only a 1.78-times higher PRR. Additional strong factors that remained in the final model for both sexes included abdominal obesity, moderate-to-severe fatty liver and a family history of GSD, which was higher in men. Only in women diabetes was strongly associated with GSD independently of abdominal obesity, while having three or more children remained a relevant factor (PRR: 1.16 (95% CI 1.09 to 1.24). Other factors associated with GSD in women but not in men included low health coverage, few remaining teeth and direct bilirubin. In both sexes, GSD was associated with a family history of gallstones, being a stronger risk factor for men (PRR 1.60, 95% CI (1.40 to 1.82)) than for women (PRR 1.23 95% CI (1.15 to 1.34) (online supplemental table 5)).
Factors associated with prevalent gallstone disease at baseline in 4838 women and 3936 men. MAUCO 2015–2019. Multiple robust Poisson regression model. GBC, gallbladder cancer; GS, gallstones; HDL, high-density lipoprotein; M, men; PRR, prevalence rate ratio; W, women.
Follow-up ultrasound findings
We invited all 2353 participants with GSD, 1811 non-GSD ultrasound-cohort controls and 595 non-GSD clinical controls for the follow-up ultrasound (online supplemental figure 1). The response rate was 89% in GSD cases, 93% in non-GSD ultrasound cohort controls and 100% in non-GSD opportunistic controls (online supplemental figure 1). During the follow-up, characteristics were similar between cases and non-GSD ultrasound cohort controls (online supplemental table 6). Findings did not change when excluding the opportunistic controls; thus, we presented the combined controls in the tables and text (online supplemental table 6).
Changes in GSD status during follow-up
Among 2239 participants with normal gallbladders at enrolment, 8.2% had GSD and 3.7% other anomalies at follow-up. Particularly notable, women had a 2.4-times higher risk of cholecystectomy (table 2A). Among 45 participants with other anomalies at baseline, 18.5% cleared their anomalies, 24.4% developed gallstones and 11.1% received a cholecystectomy; these results were similar by sex (table 2B). Among participants with gallstones since enrolment, 45.6% had a cholecystectomy, which was more likely in women (+69%, p<0.001); only nine participants (1.1%) cleared their gallstones (table 2C).
Changes in gallbladder status over 2.4 (IQR: 2.04–2.84) years of follow-up among MAUCO participants by sex 2015–2023
Associations between risk factors and GSD incidence
Over a median follow-up of 2.4 years (IQR: 2.04–2.84 years), 198 (139 women and 59 men) out of 2284 participants without GSD at baseline developed GSD.
The overall risk of GSD incidence increased by 2% (95% CI 1.005% to 1.03%) per each centimetre above the ideal waist circumference (online supplemental table 7). The association was similar in women but not in men (online supplemental table 7). Another strong predictor of GSD incidence was age (HR: 1.07 (95% CI 1.04 to 1.10)) (online supplemental table 8). When other predictors were investigated by sex, having a family history of gallbladder cancer (HR: 6.02 (95% CI 1.42 to 25.5)) and low HDL cholesterol (HR: 1.89 (95% CI 1.12 to 3.20)) were the strongest predictor in men but not in women (table 3).
Risk factors of GSD incidence by sex. MAUCO 2015–2023
Finally, the cumulative incidence of GSD and death by sex is shown in online supplemental figure 3. As it is observed, while the cumulative incidence of death was higher in men, women had a higher GSD cumulative incidence during the follow-up.
Discussion
We report the occurrence of ultrasound-detected GSD in nearly 9000 women and men from the general population of an agricultural county in Central Chile. This study confirms that the Chilean population has one of the highest burdens of GSD (30%) reported worldwide, similar to the rates found among the Pima Indians in the USA in 1970 (48.6%).8 29
GSD prevalence
Older age, higher waist circumference, moderate-to-severe fatty liver and family history of GSD were all strongly associated with prevalent GSD. The prevalence of GSD was twice as high in women (40.4%) than in men (17.1%); one of the highest sex differentials reported, with GSD being 127% higher in women. Interestingly, Sun et al reported large heterogeneity in the sex ratio by BMI in China.30 They found that among participants with normal BMI, the female-to-male (F:M) ratio of gallstone prevalence was 1.15, but among participants with BMI >25, the F:M ratio was 2.14.30 The 2016 mean BMI in Chilean women (28.3) and men (28.0) was notably higher than the mean BMI reported in China (23.6 women and 24.3 men) or in Japan (21.8 women and 23.7 men).31 In our study, obesity measured by BMI was very similar by sex; yet, abdominal obesity in women doubled that of men, being one of the main explanatory factors of the women’s excess of GSD and potentially preventable risk factors of GSD incidence. A high prevalence of GSD has been reported in Chile from autopsy reports and cholecystography studies since 1960,32 long before the epidemic of obesity currently affecting Chilean adults and children.33 A high proportion of germline variants in the ABCG8 and TRAF3 genes, which are associated with GSD and gallbladder cancer, have been reported in the Chilean population, which could explain its GSD burden.12 34
Family history of GSD and gallbladder disease were stronger risk factors for GSD among men than women (60–70% vs 25%, respectively). This finding suggests that GSD in men has a stronger genetic component than for women, while metabolic and reproductive factors play a larger role in women. For men, GSD might be a problem of older ages, while most GSD in women have occurred during their reproductive life. Of note, the fastest GSD in women occurred during the reproductive ages, reaching its peak at age 52–15 years earlier than in men, whose peak was at 65 years. This can be explained by the highest oestrogen levels in women during the reproductive years. Oestradiol has a lithogenic effect by stimulating oestrogen receptors in the liver, which disrupts bile acid metabolism, increasing bile saturation of cholesterol and inducing gallstone formation.35 As previously described by other authors, we found that women with GSD had more children, implying longer exposure to high female hormone levels.36–38 Finally, among GSD men, cardiovascular diseases appeared to be a strong risk factor. GSD and cardiovascular disease have been shown to share many common risk factors, while a recent prospective study highlighted that men with gallbladder diseases had a higher risk of a cardiovascular event.39
Cholecystectomy
At baseline, 68% of women with GSD and only 50% of men with GSD had undergone cholecystectomy. This could be explained by women having gallstones for a longer period than men, resulting in more opportunities for surgery. Women might have also been prioritised for cholecystectomy because gallbladder cancer was the number one cause of cancer death among Chilean women until 2010.40 41 Also, women with gallstones had more digestive symptoms than men, which may cause more medical visits. There is mounting evidence of the long-term effects of cholecystectomy. The gallbladder extraction profoundly alters lipid metabolism and the enterohepatic circulation of bile. The permanent flux of diluted biliary acids to the intestines results in intestinal dysbacteriosis and a pro-inflammatory intestinal state.42 Cholecystectomy also increases steatotic liver disease and elevates the risks of digestive and hepatobiliary cancers.43 44
In our study, 16% of participants who had already had cholecystectomy reported experiencing biliary colic at enrolment, suggesting that surgery did not solve the pain in these subjects. Only 17.9% of baseline biliary colic reported for all participants could be attributed to current gallstones. This lack of specificity of abdominal gallstone symptoms has also been reported by others,45 46 suggesting that gallstone prevention is preferable to surgery.
GSD incidence
In our population, the cumulative incidence of GSD was 7.2%. While women had 57.7% higher GSD at baseline, the incidence of a new GSD was only 32.8% higher in women than in men. The latter indicates that most susceptible women acquired GSD before entering the cohort, while, for men, it is a disease of older age. Similarly, the only factor significantly associated with GSD incidence in women was waist circumference; in men, the strongest factor was older age, followed by a family history of gallbladder disease and low HDL.
The main preventable risk factor for GSD incidence was abdominal obesity, similar to other populations of diverse ethnic backgrounds.35 47 We found that GSD was associated with metabolic disorders and diabetes, independent of abdominal obesity, which is consistent with cohort studies in the USA, Europe and Taiwan.48 Aune and Vatten proposed that diabetic autonomic neuropathy affects gallbladder motility, favouring bile stasis, sludge and the stone cascade.48 Recently, Cortés proposed that insulin resistance, independent of obesity, might play a causal role in GSD.49
GSD participants also had lower LDL levels than controls, similar to a previous study in the Chilean population that suggested that individuals predisposed to GSD displayed enhanced whole-body sterol clearance, a trait of ethnic groups at higher risk of GSD.50 However, we did not see an association between LDL and GSD incidence, and the literature has been inconsistent in this regard.51 52 The heterogeneity in the association of blood lipids with GSD across populations suggests confounding by gene-environment interaction.53 54
GSD and lifestyle factors
GSD participants had a lower prevalence of risky alcohol consumption. Several studies report a protective effect of alcohol against GSD,55 including prospective studies52 56 57 and a meta-analysis.58 The mechanism for this association would be a direct effect of alcohol on lipid metabolism and gallbladder motility.57 Adherence to the Mediterranean diet was low in our population, which has been reported to be protective against gallstones.59 In Chileans, legumes were suggested as a risk factor for GSD by lowering plasma cholesterol and LDL but increasing cholesterol saturation of the bile.60 This hypothesis was confirmed in a study among men in the USA, suggesting that legumes partially interrupt the enterohepatic circulation of bile acids, increasing the hepatic secretion of cholesterol.61 Chileans and American Indians consume high quantities of legumes, but prospective confirmatory studies are needed before considering a recommendation on this matter.
MAUCO is a unique Latin American cohort that enables analysis of GSD in a high-risk population with a wide range of exposures. The response rate was high, facilitated by a comprehensive surveillance system and low attrition.17 However, this study also has limitations. First, MAUCO does not represent the overall Chilean population since the study was conducted in an agricultural county in Central Chile, which also limited the evaluation of potential geographical differences that may affect the GSD development. Second, we made many comparisons, and some of our findings may be explained by chance or confounding by unmeasured or uncontrolled factors. Also, the low number of cases in some variables could determine a lack of power to identify other risk or protective factors. Nonetheless, it is reassuring that our findings agree with studies conducted in different populations worldwide. Third, the short follow-up period of this cohort may explain the lack of associations between some risk factors and GSD incidence. Therefore, analyses must be conducted again in the upcoming years or in a similar population with longer follow-ups. Finally, as per any observational study, causality cannot be inferred.
In conclusion, GSD had a different presentation by sex, occurring much earlier in women associated with female hormones, higher cholecystectomy and less impact of genetics. In men, GSD occurred at older ages, mainly in men with cholecystectomy, showing a stronger genetic component and was associated with cardiovascular risks. Women and men shared the only preventable risk factor: abdominal obesity. Therefore, improving diet quality and physical activity to decrease obesity and the associated metabolic risk factors need to be further encouraged and stressed even after cholecystectomy to diminish the risks of liver steatosis.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by ethics committees at the School of Medicine of Pontificia Universidad Católica de Chile (Number: 14-141) and the Maule Regional Public Health Service. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We would like to thank the health auxiliaries who collaborated with our study: Cintia Illanes, Loreto Ponce, Daniela Poblete, Javiera Alvarado, Karin Mancilla, Patricia Morales, Silvana Reyes, Claudia Dinamarca; data processing and statistical support: Claudia Marco, Cristián Herrera, Cristián Pinto, Fabio Paredes; ultrasound technician: Fernando Herrera; biobank support: Ian Reyes, Katherine Brito; study coordination support: Pía Venegas, Johanna Acevedo; Molina Hospital support: Miguelina Riquelme, Nilton Palacios; Molina major: Priscila Castillo; Maule Health Service collaborator: Roberto Bachler. We especially thank all the participants in this study who generously gave their time and provided us with valuable information and biospecimens.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors DR: data analysis, data interpretation, manuscript drafting, critical revision. LH, VVdW, MG: data acquisition, data interpretation. FG, VC: data processing, data analysis, manuscript drafting. CV: data analysis, data interpretation, critical revision. FP-R: data interpretation, critical revision. FC, PC: data acquisition, critical revision. JCR, JCA, JK, SC: critical revision. MA: study design, data interpretation, critical revision. CF (article guarantor): study conception and design, data analysis, data interpretation, manuscript drafting, critical revision. All authors have approved the submitted manuscript.
Funding This work was supported by grants from Comisión Nacional de Investigación Científica y Tecnológica (CONICYT), a Chilean governmental scientific commission: Fondo Nacional de Desarrollo Científico y Tecnológico to CF (grant number 1170395) and to MA (grant number 1191145); Fondo de Financiamiento de Centros de Investigación en Áreas Prioritarias (grant number 15130011); and CONICYT PIA/Basal PFB12, Basal Centre for Excellence in Science and Technology to MA. We also received material support from the Faculty of Medicine of Universidad Católica del Maule; the School of Medicine of the Pontificia Universidad Católica de Chile; the Ilustre Municipalidad de Molina (Molina’s municipality); and Molina and Curicó hospitals. Additional funds come from the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Division of Cancer Epidemiology and Genetics and the Office of Research on Women's Health of the United States National Institutes of Health.
Disclaimer The funding bodies had no role in the study design, the collection, analysis and interpretation of data, or the writing of the article and the decision to submit it for publication.
Competing interests None declared.
Provenance and peer review Not commissioned; internally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.