Lifestyle and dietary factors associated with serologically detected gastric atrophy in a Caucasian population in the GISTAR study

OBJECTIVE

To identify dietary and lifestyle factors associated with decreased pepsinogen levels indicative of gastric atrophy.

METHODS

Participants aged 40 to 64 from the "Multicentric randomized study of H. pylori eradication and pepsinogen testing for prevention of gastric cancer mortality (GISTAR study)" in Latvia tested for serum pepsinogen, as well as for Helicobacter pylori infection by 13 C-urea breath test or serology were included. Data on sex, age, education, employment, diet, smoking, alcohol and proton pump inhibitor use were obtained by survey and compared for participants with and without serologically detected gastric atrophy defined as pepsinogen I/pepsinogen II ≤ 2 and pepsinogen I ≤ 30 ng/mL.

RESULTS

Of 3001 participants (median age 53, interquartile range, 11.0, 36.9% male) 52.8% had H. pylori and 7.7% had serologically detected gastric atrophy. In multivariate analysis, increasing age, consumption of alcohol, coffee, and onions were positively, while H. pylori , former smoking, pickled product and proton pump inhibitor use were inversely associated with gastric atrophy. Pepsinogen values were higher in smokers and those with H. pylori . Pepsinogen ratio was lower in those with H. pylori . When stratifying by H. pylori presence, significantly higher pepsinogen levels remained for smokers without H. pylori .

CONCLUSION

Several dietary factors and smoking were associated with serologically detected gastric atrophy. Pepsinogen levels differed by smoking and H. pylori status, which may affect the serologic detection of gastric atrophy. There seems to be a complicated interaction between multiple factors. A prospective study including atrophy determined by both serology and histology is necessary.

Assessment of Serum Pepsinogens with and without Co-Testing with Gastrin-17 in Gastric Cancer Risk Assessment-Results from the GISTAR Pilot Study

Introduction−−Serum pepsinogen tests for gastric cancer screening have been debated for decades. We assessed the performance of two pepsinogen assays with or without gastrin-17 for the detection of different precancerous lesions alone or as a composite endpoint in a Latvian cohort. Methods−−Within the intervention arm of the GISTAR population-based study, participants with abnormal pepsinogen values by ELISA or latex-agglutination tests, or abnormal gastrin-17 by ELISA and a subset of subjects with all normal biomarker values were referred for upper endoscopy with biopsies. Performance of biomarkers, corrected by verification bias, to detect five composite outcomes based on atrophy, intestinal metaplasia, dysplasia or cancer was explored. Results−−Data from 1045 subjects were analysed, of those 273 with normal biomarker results. Both pepsinogen assays showed high specificity (>93%) but poor sensitivity (range: 18.4−31.1%) that slightly improved when lesions were restricted to corpus location (40.5%) but decreased when dysplasia and prevalent cancer cases were included (23.8%). Adding gastrin-17 detection, sensitivity reached 33−45% while specificity decreased (range: 61.1−62%) and referral rate for upper endoscopy increased to 38.6%. Conclusions−−Low sensitivity of pepsinogen assays is a limiting factor for their use in population-based primary gastric cancer screening, however their high specificity could be useful for triage.

Factors Associated with False Negative Results in Serum Pepsinogen Testing for Precancerous Gastric Lesions in a European Population in the GISTAR Study

The accuracy of plasma pepsinogen (Pg) as a marker for precancerous gastric lesions (PGL) has shown variable results. We aimed to identify factors associated with false negative (FN) cases in Pg testing and to adjust cut-off values for these factors in order to improve Pg yield. Plasma Pg was measured and upper endoscopy with biopsy was performed within the "Multicentric randomized study of Helicobacter pylori eradication and pepsinogen testing for prevention of gastric cancer mortality: the GISTAR study". A multivariable logistic model was built for FN and multiple factors. Values of Pg were compared and sensitivity and specificity were calculated using pre-existing Pg cut-offs for factors showing strong associations with FN. New cut-offs were calculated for factors that showed substantially lower sensitivity. Of 1210 participants, 364 (30.1%) had histologically confirmed PGL, of which 160 (44.0%) were FN. Current smokers, men, and H. pylori positives were more likely FN. Smoking in H. pylori negatives was associated with a higher Pg I/II ratio and substantially lower sensitivity of Pg testing than in other groups. Adjusting Pg cut-offs for current smokers by H. pylori presence improved sensitivity for detecting PGL in this group. Our study suggests that adjusting Pg cut-offs for current smokers by H. pylori status could improve Pg test performance.

Taxonomy-based data representation for data mining: an example of the magnitude of risk associated with H. pylori infection

BACKGROUND

The amount of available and potentially significant data describing study subjects is ever growing with the introduction and integration of different registries and data banks. The single specific attribute of these data are not always necessary; more often, membership to a specific group (e.g. diet, social 'bubble', living area) is enough to build a successful machine learning or data mining model without overfitting it. Therefore, in this article we propose an approach to building taxonomies using clustering to replace detailed data from large heterogenous data sets from different sources, while improving interpretability. We used the GISTAR study data base that holds exhaustive self-assessment questionnaire data to demonstrate this approach in the task of differentiating between H. pylori positive and negative study participants, and assessing their potential risk factors. We have compared the results of taxonomy-based classification to the results of classification using raw data.

RESULTS

Evaluation of our approach was carried out using 6 classification algorithms that induce rule-based or tree-based classifiers. The taxonomy-based classification results show no significant loss in information, with similar and up to 2.5% better classification accuracy. Information held by 10 and more attributes can be replaced by one attribute demonstrating membership to a cluster in a hierarchy at a specific cut. The clusters created this way can be easily interpreted by researchers (doctors, epidemiologists) and describe the co-occurring features in the group, which is significant for the specific task.

CONCLUSIONS

While there are always features and measurements that must be used in data analysis as they are, the use of taxonomies for the description of study subjects in parallel allows using membership to specific naturally occurring groups and their impact on an outcome. This can decrease the risk of overfitting (picking attributes and values specific to the training set without explaining the underlying conditions), improve the accuracy of the models, and improve privacy protection of study participants by decreasing the amount of specific information used to identify the individual.

Does Family History of Cancer Influence Undergoing Screening and Gastrointestinal Investigations?

BACKGROUND AND AIMS

Although a family history of cancer (FHC) can modify the lifestyle and attitudes towards participation in cancer screening programs, studies on this relationship show mixed results and vary across populations. The objectives of the study were to compare sociodemographic characteristics, history of gastrointestinal (GI) investigations and Helicobacter pylori eradication, and modifiable cancer risk factors between those with FHC and those with no FHC (NFHC), and to investigate the association between FHC and a history of GI investigations.

METHODS

A total of 3,455 questionnaires from the pilot study of the "Helicobacter pylori eradication and pepsinogen testing for prevention of gastric cancer mortality (GISTAR study)" in Latvia were analysed. We compared sociodemographic characteristics and history of GI investigations between participants with self- reported FHC and NFHC. Binary logistic regression models adjusted for socio-demographic characteristics and modifiable cancer risk factors were built for a FHC and each GI investigation.

RESULTS

Participants with a FHC were more likely to be women, have a higher education and less likely to have harmful habits (smoking, alcohol consumption) than those with NFHC. Participants with a FHC were approximately twice as likely to report recent colorectal investigations specifically for screening, than those with NFHC. In fully adjusted logistic regression models, FHC was significantly associated with a recent history of faecal occult blood tests (FOBTs), colonoscopies, and colorectal investigations (FOBT or colonoscopy) specifically for screening as part of the national organized screening programme.

CONCLUSION

Our results indicate that those with a FHC have different patterns of health-related behaviour than those with NFHC.

Sociodemographic, Lifestyle and Medical Factors Associated with Helicobacter Pylori Infection

BACKGROUND AND AIMS

The prevalence of Helicobacter pylori (H. pylori) infection is higher in developing countries and is often linked to lower socioeconomic status. Few studies have investigated the association between H. pylori and individual level characteristics in Europe, where several countries have a high prevalence of H. pylori infection. The study aimed to identify risk factors for H. pylori infection among adults in a large clinical trial in Latvia.

METHODS

1,855 participants (40-64 years) of the "Multicenter randomized study of H. pylori eradication and pepsinogen testing for prevention of gastric cancer mortality" (GISTAR study) in Latvia tested for H. pylori IgG antibodies were included in a cross-sectional analysis. Sociodemographic, lifestyle and medical factors were compared for participants seropositive (H. pylori+) and seronegative. Mutually adjusted odds ratios (OR) were calculated for H. pylori+ and factors significant in univariate analysis (education, smoking, binge drinking, several dietary habits, history of H. pylori eradication and disease), adjusting for age, gender and income.

RESULTS

Of the participants 1,044 (55.4%) were H. pylori seropositive. The infection was associated with current (OR: 1.34, 95%CI: 1.01-1.78) and former (OR: 1.38; 95%CI: 1.03-1.85) smoking, binge drinking (OR: 1.35; 95%CI: 1.03-1.78), having ≥200g dairy daily (OR: 1.37; 95%CI: 1.11-1.69), and very hot food/drinks (OR: 1.32; 95%CI: 1.03-1.69) and inversely with ≥400g vegetables/fruit daily (OR: 0.76; 95%CI: 0.60-0.96), history of H. pylori eradication (OR: 0.57; 95%CI: 0.39-0.84), peptic ulcer (OR: 0.55; 95%CI: 0.38-0.80) and cardiovascular disease (OR: 0.78; 95%CI: 0.61-0.99).

CONCLUSIONS

After mutual adjustment, H. pylori seropositivity was associated with lifestyle and in particular dietary factors rather than socioeconomic indicators in contrast to the majority of other studies.