<![CDATA[Background: Gastro-oesophageal reflux disease is among the most prevalent chronic gastrointestinal disorders worldwide and its absolute burden continues to rise, yet studies of behavioural determinants in young adults report statistical significance without magnitude and almost never state the effect size their design could detect. Objective: To estimate the association of dietary pattern and psychological stress with symptom-defined gastro-oesophageal reflux, and the precision with which those associations could be detected, in a restriction-based sample of female preclinical medical students. Methods: This STROBE-compliant cross-sectional study used a restriction-based design to isolate behavioural exposures from adiposity and the other dominant reflux risk factors. Of 120 randomly selected female preclinical medical students at the Faculty of Medicine, Universitas Pembangunan Nasional "Veteran" Jakarta, 104 responded (86.7%) and 87 met eligibility criteria fixing female sex, body mass index 18.5-24.9 kg/m2, never-smoking, no alcohol intake and no gastro-active medication. Dietary pattern, stress and reflux symptoms were measured with validated Indonesian instruments and analysed with exact tests, Firth penalised logistic regression, Bayesian estimation, quantitative bias analysis and formal precision analysis. Results: GERD-Q positivity was 12.6% (95% CI 7.2-21.2); 8.0% had a poor dietary pattern and 43.7% reported stress. Adjusted odds ratios were 3.35 (95% CI 0.55-16.12; p = 0.208) for poor dietary pattern and 1.10 (95% CI 0.31-3.78; p = 0.339) for stress; Bayes factors favouring the null ranged from 0.99 to 2.05 and from 1.78 to 4.90 across four priors. Minimum detectable odds ratios were 12.23 and 4.32. Correcting for outcome misclassification raised the dietary odds ratio to 6.61. Conclusion: Neither exposure was associated with reflux symptoms; the study excludes only large effects, and about 459 participants would be needed to detect the observed dietary effect.]]>
