Findings of diagnostic colonoscopy in young adults versus findings of screening colonoscopy in patients aged 50 to 54 years: a comparative study stratified by symptom category

Background: The threshold for diagnostic colonoscopy in symptomatic patients aged <50 years remains controversial. Previous studies on the prevalence of neoplasia or other serious pathology in young patients mostly have been uncontrolled, providing only limited data on the risk associated with specific symptoms. Objective: To compare colonoscopy findings in patients aged <50 years who have various symptoms (diagnostic cohort) against those of concurrent patients aged 50 to 54 years who are asymptomatic (screening cohort). Design: Retrospective controlled cohort study. Setting: Teaching hospital. Patients: Symptomatic patients aged between 18 and 49 years and asymptomatic patients aged between 50 and 54 years. Interventions: Colonoscopy. Main Outcome Measurements: Prevalence of advanced neoplasia. Results: During the study period, 1638 patients underwent colonoscopy in the screening cohort (mean [+/- standard deviation{SD}] age 51.7 +/- 1.4 years) and 1266 underwent colonoscopy in the diagnostic cohort (40.4 +/- 8.0 years). Despite the age difference, the prevalence of advanced neoplasia in patients with rectal bleeding was comparable with that in the screening controls: 28 of 472 (5.9%) versus 113 of 1638 patients (6.9%) (P = .459). Furthermore, 10 patients (2.1%) with rectalbleeding were newly diagnosed with inflammatory bowel disease. In contrast, other symptoms that commonly lead to colonoscopy, such as abdominal pain, changes in bowel habits, and weight loss, were associated with much lower risks for neoplasia. As a result, the overall prevalences of neoplasia and advanced neoplasia were significantly higher in the screening cohort than in the diagnostic cohort: 467 of 1638 patients (28.5%) versus 179 of 1266 patients (14.1%), and 113 patients (6.9%) versus 48 patients (3.8%), respectively (both P < .001). Limitations: No data on duration of symptoms; discrepant sex ratios between cohorts. Conclusion: The threshold for diagnostic colonoscopy in symptomatic young adults should be individualized for each symptom category. Rectal bleeding warrants colonoscopy to detect advanced neoplasia or inflammatory bowel disease in most young patients, especially those aged 40 to 49 years, whereas non-bleeding symptoms, including some traditionally regarded as "alarm" symptoms, were associated with a much lower risk for neoplasia compared with the risk in screening patients aged 50 to 54 years.