Age thresholds have long guided endoscopic evaluation for dyspepsia, yet their clinical validity in contemporary low-risk populations—and their applicability to gastroesophageal reflux disease (GERD)—remain uncertain. With declining gastric cancer incidence and growing emphasis on risk-based decision models, re-examining the diagnostic value of age is timely. This study aimed to assess whether age effectively predicts clinically significant findings (CSFs) in patients with treatment-refractory dyspepsia or GERD and to define optimal age cutoffs for endoscopic triage.

We retrospectively analyzed a large consecutive cohort of adults undergoing diagnostic upper endoscopy for dyspepsia or GERD following inadequate response to proton pump inhibitor (PPI) therapy. Individuals with alarm features, prior upper gastrointestinal malignancy, or recent endoscopy were excluded. Endoscopic findings were classified as CSFs if they carried diagnostic or therapeutic relevance, including Barrett’s esophagus, severe esophagitis, peptic ulcer disease, or neoplasia. Logistic regression was used to identify independent predictors of CSFs, and receiver operating characteristic (ROC) analysis was applied to evaluate the discriminative capacity of age and determine optimal cutoffs.

Among 20,248 procedures (13,842 for dyspepsia and 6,406 for GERD), endoscopic yield differed markedly by indication. CSFs were detected in 12.8% of GERD and 5.6% of dyspepsia cases. Esophageal pathology—particularly erosive esophagitis, Barrett’s esophagus, and hiatal hernia—dominated among reflux patients, whereas gastric and duodenal lesions were more frequent in dyspepsia. Multivariate analysis revealed that age ≥60 years independently predicted CSFs in GERD (adjusted OR 1.66, 95% CI 1.38–2.00) but not in dyspepsia (OR 0.97, 95% CI 0.82–1.15). Male sex was a consistent predictor across both indications, and in GERD, Arab ethnicity and hiatal hernia also conferred increased risk. Malignancy was exceedingly rare (<0.15%). ROC analysis demonstrated minimal discriminative power for age (AUC 0.574 in GERD; 0.512 in dyspepsia), with Youden-derived optimal thresholds of 59 and 52 years, respectively—offering limited clinical utility.

In this large real-world cohort of non-alarm, PPI-refractory patients, age alone was a weak discriminator of clinically significant endoscopic pathology. Although older age modestly increased diagnostic yield in GERD, its overall predictive performance was poor, particularly in dyspepsia. These findings challenge the continued reliance on uniform age thresholds for endoscopy referral and support the adoption of context-specific, multifactorial risk models that integrate demographic and structural predictors to optimize diagnostic efficiency and resource utilization.