Dataset Information

Predictive Factors for Future Onset of Reflux Esophagitis: A Longitudinal Case-control Study Using Health Checkup Records.

ABSTRACT:

Background/aims

Although risk factors of reflux esophagitis (RE) have been investigated in numerous cross-sectional studies, little is known about predictive factors associated with future onset of RE. We investigated time courses of clinical parameters before RE onset by a longitudinal case-control study using health checkup records.

Methods

We used health checkup records between April 2004 and March 2014 at 9 institutions in Japan. A multivariate logistic regression analysis was performed to evaluate associations of baseline clinical parameters with RE. The time courses of the clinical parameters of RE subjects were compared with those of non-RE subjects by the mixed-effects models for repeated measures analysis or longitudinal multivariate logistic analysis.

Results

Initial data were obtained from 230 056 individuals, and 2066 RE subjects and 4132 non-RE subjects were finally included in the analysis. Body mass index, alanine aminotransferase, smoking, acid reflux symptoms, hiatal hernia, and absence of atrophic gastritis at baseline were independently associated with RE. The time courses of body mass index, fasting blood sugar, triglyceride, aspartate aminotransferase, alanine aminotransferase, γ-glutamyl transpeptidase, percentages of acid reflux symptoms, feeling of fullness, and hiatal hernia in the RE group were significantly worse than in the non-RE group.

Conclusions

The RE group displayed a greater worsening of the clinical parameters associated with lifestyle diseases, including obesity, diabetes, hyperlipidemia, and fatty liver for 5 years before RE onset compared with the non-RE group. These results suggest that RE is a lifestyle disease and thus lifestyle guidance to at-risk person may help to prevent RE onset.

SUBMITTER: Toki Y

PROVIDER: S-EPMC8748855 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Background and study aims ?Non-erosive reflux disease (NERD) includes minimal change esophagitis (MCE) and no endoscopic abnormalities. However, for most endoscopists, it is difficult to detect MCE with conventional white-light endoscopy (WLE). Linked color imaging (LCI) technology is the most recently developed image-enhancing technology and improves detection and differentiation of subtle mucosal changes using a color contrast method. This study assessed the efficacy of WLE combined with LCI for diagnosing MCE compared with WLE. Patients and methods ?Between February and May 2017, 44 NERD patients and 40 healthy subjects were enrolled in our study. First, the distal esophagus was examined using WLE followed by LCI. Second, three experienced endoscopists observed all the patients' white-light (WL) images and corresponding images of WL and LCI and then recorded presence or absence of minimal change esophagitis (MCE?+/-). The proportion of minimal change between the two groups was then compared. Third, five blinded endoscopists with different levels of endoscopic experience assessed whether MCE was present. Intraobserver reproducibility and interobserver agreement were described using the kappa value. Results ?The proportion of MCE in the NERD group (70.8?%, 35/48) was higher than that in the control group (22.5?%, 9/40, P ?<?0.001) when diagnosed by the three experienced endoscopists. Detection rates for MCE using WLE combined with LCI were higher than those using WLE (43/88, 48.9?% vs. 29/88, 33.0?%, P ?<?0.001). With WLE combined with LCI, intraobserver reproducibility signi?cantly improved, indicating that the combined approach can improve interobserver agreement compared with using WLE alone. Conclusions ?Endoscopic diagnosis of MCE using WLE combined with LCI images is effective. Intraobserver reproducibility and interobserver agreement in MCE can be improved when LCI is applied with conventional imaging (Clinical trial registration number: NCT03068572).

Project description:Abstract Objective Hypertension has long been recognized as one of the most important predisposing factors for cardiovascular diseases and mortality. In recent years, machine learning methods have shown potential in diagnostic and predictive approaches in chronic diseases. Electronic health records (EHRs) have emerged as a reliable source of longitudinal data. The aim of this study is to predict the onset of hypertension using modern deep learning (DL) architectures, specifically long short-term memory (LSTM) networks, and longitudinal EHRs. Materials and Methods We compare this approach to the best performing models reported from previous works, particularly XGboost, applied to aggregated features. Our work is based on data from 233 895 adult patients from a large health system in the United States. We divided our population into 2 distinct longitudinal datasets based on the diagnosis date. To ensure generalization to unseen data, we trained our models on the first dataset (dataset A “train and validation”) using cross-validation, and then applied the models to a second dataset (dataset B “test”) to assess their performance. We also experimented with 2 different time-windows before the onset of hypertension and evaluated the impact on model performance. Results With the LSTM network, we were able to achieve an area under the receiver operating characteristic curve value of 0.98 in the “train and validation” dataset A and 0.94 in the “test” dataset B for a prediction time window of 1 year. Lipid disorders, type 2 diabetes, and renal disorders are found to be associated with incident hypertension. Conclusion These findings show that DL models based on temporal EHR data can improve the identification of patients at high risk of hypertension and corresponding driving factors. In the long term, this work may support identifying individuals who are at high risk for developing hypertension and facilitate earlier intervention to prevent the future development of hypertension.

Project description:BACKGROUND:It is difficult to obtain detailed information on non-participants in physical and health examination checkups in community-based epidemiological studies. We investigated the characteristics of non-participants in a physical and health examination checkup for older adults in a nested study from the Japanese Kyoto-Kameoka Longitudinal Study. METHODS:We approached a total of 4831 people aged ?65 years in 10 randomly selected intervention regions. Participants responded to a mail-based population survey on needs in the sphere of daily life to encourage participation in a free face-to-face physical checkup examination; 1463 participants (706 men, 757 women) participated in the physical checkup. A multiple logistic regression model was performed to investigate the adjusted odds ratios (aOR) of non-participation based on sociodemographic status apart from psychological and physiological frailty as assessed by the validated Kihon Checklist. RESULTS:There was a significant, inverse relationship between non-participation and frequently spending time alone among individuals who lived with someone or other family structure (aOR?=?0.53, standard error [SE] 0.08 in men, aOR?=?0.66, SE 0.09 in women). Very elderly (over 80 years old) women, poorer health consciousness and current smoking in both sexes and poor self-rated health in men, were significantly related to higher non-participation rates. In both sexes, individuals who did not participate in community activities were significantly more likely to be non-participants than individuals who did (aOR?=?1.94, SE 0.23 in men, aOR?=?3.29, SE 0.39 in women). Having low IADL and physical functioning scores were also associated with higher rates of non-participation. CONCLUSION:Health consciousness and lack of community activity participation were predictors of non-participation in a physical checkup examination among older adults. In addition, lower IADL and physical functioning/strength were also predictors of non-participation. On the contrary, older inhabitants living with someone tended to participate in the physical checkup examination for social interchange when they were frequently alone in the household. This study suggests the importance of considering aging especially for women and poor sociodemographic background and physical frailty for both sexes so that older people can access health programs without difficulty. TRIAL REGISTRATION:UMIN000008105 . Registered 26 April 2012. Retrospectively registered.

Project description:BackgroundDespite the high prevalence of gastroesophageal reflux disease (GERD), its risk factors are still a subject of controversy. This is probably due to inadequate distinction between reflux esophagitis (RE) and non-erosive reflux disease (NERD), and is also due to inadequate evaluation of adjacent stomach. Our aim is therefore to define background factors of RE and NERD independently, based on the evaluation of Helicobacter pylori infection and gastric atrophy.MethodsWe analyzed 10,837 healthy Japanese subjects (6,332 men and 4,505 women, aged 20-87 years) who underwent upper gastrointestinal endoscopy. RE was diagnosed as the presence of mucosal break, and NERD was diagnosed as the presence of heartburn and/or acid regurgitation in RE-free subjects. Using GERD-free subjects as control, background factors for RE and NERD were separately analyzed using logistic regression to evaluate standardized coefficients (SC), odds ratio (OR), and p-value.ResultsOf the 10,837 study subjects, we diagnosed 733 (6.8%) as RE and 1,722 (15.9%) as NERD. For RE, male gender (SC = 0.557, OR = 1.75), HP non-infection (SC = 0.552, OR = 1.74), higher pepsinogen I/II ratio (SC = 0.496, OR = 1.64), higher BMI (SC = 0.464, OR = 1.60), alcohol drinking (SC = 0.161, OR = 1.17), older age (SC = 0.148, OR = 1.16), and smoking (SC = 0.129, OR = 1.14) are positively correlated factors. For NERD, HP infection (SC = 0.106, OR = 1.11), female gender (SC = 0.099, OR = 1.10), younger age (SC = 0.099, OR = 1.10), higher pepsinogen I/II ratio (SC = 0.099, OR = 1.10), smoking (SC = 0.080, OR = 1.08), higher BMI (SC = 0.078, OR = 1.08), and alcohol drinking (SC = 0.076, OR = 1.08) are positively correlated factors. Prevalence of RE in subjects with chronic HP infection and successful HP eradication denotes significant difference (2.3% and 8.8%; p<0.0001), whereas that of NERD shows no difference (18.2% and 20.8%; p = 0.064).ConclusionsSignificantly associated factors of NERD are considerably different from those of RE, indicating that these two disorders are pathophysiologically distinct. Eradication of Helicobacter pylori may have disadvantageous effects on RE but not on NERD.