To the Editor: We conducted a study to compare the effects of, and any correlation between, chronic nonsteroidal antiinflammatory agent (NSAID) intake and Helicobacter pylori (HP) infection, in terms of upper gastrointestinal (GI) symptoms and lesions, in younger compared with older subjects. NSAIDs and HP are known to interfere with gastroduodenal mucosal protective mechanisms, and are implicated in the etiology of peptic ulcer disease.1 In older people, NSAID usage is associated with more GI complications,2 and HP infection is also more common.3 Controversy exists in the relationship between NSAID use and HP infection in peptic ulcer disease: while some studies found NSAID-related ulcers to be more common in HP-infected subjects,4 other researchers have shown that although HP prevalence increases with age, this was independent of NSAID intake and showed no correlation with the severity of dyspeptic symptoms.5,6 A total of 186 subjects referred for upper GI endoscopy for suspected GI hemorrhage, dyspepsia, or anemia were recruited from medical clinics and wards. Exclusion criteria included use of warfarin or steroids in the past year, previous upper GI surgery, actively bleeding lesions, and conditions in which endoscopy or biopsy were contraindicated. Subjects were stratified according to age (<50 or >60 years), NSAID intake (therapeutic doses for >3 months), and HP infection status. All subjects were interviewed for demographic data, cigarette and alcohol consumption, and details of NSAID or anti-ulcer drug intake, if any, and past history of peptic ulcer disease (Table 1). Frequency and severity of upper GI symptoms, including dyspepsia, pain, bloating, and gas, were graded by a standardized score. At endoscopy, lesions were recorded by the modified Lanza score,7 and antral biopsies were obtained for the CLO test,8 histology, and culture to confirm HP infection. Our findings showed that the HP infection rate was 42% in subjects aged less than 50 and 49% in those over age 65. Among NSAID users, there was a nonsignificant trend (P = .29) toward a lower HP prevalence in younger subjects (35 %) compared with the older group (50%). Table 2 summarizes the relationship between NSAID use, HP infection, upper GI symptoms, and endoscopic findings. Overall, significantly more older subjects were asymptomatic, and NSAID users were less symptomatic at all ages compared with non-users. Younger non-users had more severe symptoms, particularly those who were infected with HP. In terms of endoscopic lesions, there was no correlation with NSAID intake in the younger group. On the other hand, older chronic NSAID users were less likely to have normal endoscopic examinations, particularly the subgroup who were infected with HP. This subgroup also had significantly more mucosal hemorrhages (44%) compared with non-NSAID users (12%) and showed a trend toward more duodenal and gastric ulcers. Among young NSAID users, there was significantly more HP infection found in those with duodenal ulcers than in uninfected individuals. The lower than predicted HP infection rate in our older subjects may be related to the high prevalence of infection at an earlier age (childhood or adolescence) among Orientals.9 Our study has shown that in older people, dyspeptic symptoms are less prominent, particularly in subjects infected with HP, although they were more likely to have abnormal endoscopic findings. This highlights the need for caution in the management of upper GI problems in this group of patients. In summary, our findings suggest that in older subjects, HP infection is associated with more upper GI pathology, and chronic NSAID intake has an independent and additive effect to HP in causing such injuries, although symptoms are few in this group.