Risk Factors for Barrett Esophagus
Risk Factors for Barrett Esophagus
We analyzed data on 1,952 patients (35 AA cases, 266 NHW cases, 669 AA controls, and 982 NHW controls); their mean age was 60.3 years (s.d. 8.1) and 91.5% were men. Prevalence of BE was significantly higher among NHWs (21.3%; 95% CI, 19.1–23.7%) compared with that among AAs (5.0%; 95% CI, 3.5–6.9%; P<0.001). Excluding BE cases diagnosed before the study period, the incidence of BE remained significantly higher among NHWs (15.1%; 95% CI, 13.1–17.3%) vs. AAs (4.2%; 95% CI, 2.9–6.0%; P<0.001).
Among all participants, NHWs were more likely to be male (P=0.002), older (P<0.001), current smokers (P=0.03), overweight or obese (P=0.006), have a high WHR (P<0.001), hiatal hernia (P<0.001), and report GERD symptoms (P<0.001), PPI use (P<0.001), and NSAID use (P=0.003) compared with AAs (Table 1 and Table 2). AAs were more likely than NHWs to have active H. pylori infection, active gastritis of the antrum and corpus, and chronic gastritis of the antrum and corpus (all P<0.001).
Among AAs, cases were statistically significantly more likely to be older, have a hiatus hernia, and report GERD symptoms and PPI use (Table 1 and Table 2). There were no significant differences in active gastritis, chronic gastritis, or WHR between AA cases and AA controls. After mutual adjustment (Table 3), only hiatus hernia ≥3 cm (OR, 4.12; 95% CI, 1.57–10.81) and GERD or PPI (OR, 3.70; 95% CI, 1.40–9.78) remained significantly associated with BE in AAs.
Among NHWs, cases were significantly more likely to be male, have a high WHR, a hiatus hernia ≥3 cm, and report GERD symptoms and PPI use than controls (Table 1 and Table 2). NHW controls were more likely to use statins than NHW cases (P=0.02). There was no significant difference in H. pylori infection among NHW cases and NHW controls. In the multivariate analysis among NHWs, high WHR (OR, 2.82, 95% CI, 1.41–5.63), hiatus hernia <3 cm (OR, 2.45; 95% CI, 1.53–3.93) and ≥3 cm (OR, 4.95; 95% CI, 3.05–8.03), and PPI use (OR, 1.88; 95% CI, 1.33–2.66) were significantly associated with an increased risk of BE (Table 3).
Among BE cases, NHWs were more likely to have a high WHR compared with AAs (P<0.001). AA cases were more likely to have H. pylori (P=0.002), active gastritis of the antrum (P=0.001) and corpus (P<0.001), and chronic gastritis of the antrum (P<0.001) and corpus (P<0.001) compared with the NHW cases. NHW cases had a significantly longer mean length of BE by 1.5 cm (P=0.003; Table 3) and a greater proportion of LSBE (P=0.002) than did AA cases. NHW cases trended toward having more cases of dysplasia than AA cases, as all AA cases had nondysplastic BE (P=0.06; Table 4). There were no differences in GERD symptoms, PPI use, and hiatus hernia among NHW cases and AA cases on univariate analysis.
Excluding cases that were positive for H. pylori infection, the prevalence of dysplasia was not significantly different between NHWs and AAs (P=0.51). However, in this subset of cases, NHW cases had a greater proportion of LSBE than AA cases (42.9% of NHW cases vs. 19.0% of AA cases; P=0.033).
In multivariate analysis of BE cases, only WHR, H. pylori infection, and chronic gastritis were retained in the model (Table 3). AA cases were less likely to have a high WHR (OR, 0.11; 95% CI, 0.04–0.31) compared with NHW cases. On the other hand, AA cases were more likely to have H. pylori (OR, 2.33; 95% CI, 0.98–4.98) and chronic gastritis (OR, 3.29; 95% CI, 1.11–9.76) than NHW cases.
Cases were less likely than controls to be AA (unadjusted OR, 0.19; 95% CI, 0.13–0.28; Table 5). In univariate analysis of all cases and controls, cases were also more likely to be male and older, have a high WHR or hiatus hernia, and report GERD symptoms ≥10 years or PPI use. Cases were less likely to have H. pylori infection, active gastritis of the corpus, or chronic gastritis of the antrum and corpus than controls (Table 5). In the full multivariate model of all cases and controls, race, sex, H. pylori infection, hiatus hernia, active gastritis in the antrum, PPI use, and NSAID use were retained in the model (Table 5). Despite adjusting for risk factors, the effect of race as a predictor of BE was not attenuated compared with the unadjusted estimates, as the risk of BE was significantly lower in AAs compared with NHWs (adjusted OR 0.26; 95% CI, 0.17–0.39). Male sex, high WHR, hiatus hernia, and active gastritis of the antrum were associated with increased odds of BE. H. pylori infection was protective against the development of BE (adjusted OR 0.60; 95% CI, 0.38–0.97).
Results
We analyzed data on 1,952 patients (35 AA cases, 266 NHW cases, 669 AA controls, and 982 NHW controls); their mean age was 60.3 years (s.d. 8.1) and 91.5% were men. Prevalence of BE was significantly higher among NHWs (21.3%; 95% CI, 19.1–23.7%) compared with that among AAs (5.0%; 95% CI, 3.5–6.9%; P<0.001). Excluding BE cases diagnosed before the study period, the incidence of BE remained significantly higher among NHWs (15.1%; 95% CI, 13.1–17.3%) vs. AAs (4.2%; 95% CI, 2.9–6.0%; P<0.001).
Among all participants, NHWs were more likely to be male (P=0.002), older (P<0.001), current smokers (P=0.03), overweight or obese (P=0.006), have a high WHR (P<0.001), hiatal hernia (P<0.001), and report GERD symptoms (P<0.001), PPI use (P<0.001), and NSAID use (P=0.003) compared with AAs (Table 1 and Table 2). AAs were more likely than NHWs to have active H. pylori infection, active gastritis of the antrum and corpus, and chronic gastritis of the antrum and corpus (all P<0.001).
AA Cases vs. AA Controls
Among AAs, cases were statistically significantly more likely to be older, have a hiatus hernia, and report GERD symptoms and PPI use (Table 1 and Table 2). There were no significant differences in active gastritis, chronic gastritis, or WHR between AA cases and AA controls. After mutual adjustment (Table 3), only hiatus hernia ≥3 cm (OR, 4.12; 95% CI, 1.57–10.81) and GERD or PPI (OR, 3.70; 95% CI, 1.40–9.78) remained significantly associated with BE in AAs.
NHW Cases vs. NHW Controls
Among NHWs, cases were significantly more likely to be male, have a high WHR, a hiatus hernia ≥3 cm, and report GERD symptoms and PPI use than controls (Table 1 and Table 2). NHW controls were more likely to use statins than NHW cases (P=0.02). There was no significant difference in H. pylori infection among NHW cases and NHW controls. In the multivariate analysis among NHWs, high WHR (OR, 2.82, 95% CI, 1.41–5.63), hiatus hernia <3 cm (OR, 2.45; 95% CI, 1.53–3.93) and ≥3 cm (OR, 4.95; 95% CI, 3.05–8.03), and PPI use (OR, 1.88; 95% CI, 1.33–2.66) were significantly associated with an increased risk of BE (Table 3).
NHW Cases vs. AA Cases
Among BE cases, NHWs were more likely to have a high WHR compared with AAs (P<0.001). AA cases were more likely to have H. pylori (P=0.002), active gastritis of the antrum (P=0.001) and corpus (P<0.001), and chronic gastritis of the antrum (P<0.001) and corpus (P<0.001) compared with the NHW cases. NHW cases had a significantly longer mean length of BE by 1.5 cm (P=0.003; Table 3) and a greater proportion of LSBE (P=0.002) than did AA cases. NHW cases trended toward having more cases of dysplasia than AA cases, as all AA cases had nondysplastic BE (P=0.06; Table 4). There were no differences in GERD symptoms, PPI use, and hiatus hernia among NHW cases and AA cases on univariate analysis.
Excluding cases that were positive for H. pylori infection, the prevalence of dysplasia was not significantly different between NHWs and AAs (P=0.51). However, in this subset of cases, NHW cases had a greater proportion of LSBE than AA cases (42.9% of NHW cases vs. 19.0% of AA cases; P=0.033).
In multivariate analysis of BE cases, only WHR, H. pylori infection, and chronic gastritis were retained in the model (Table 3). AA cases were less likely to have a high WHR (OR, 0.11; 95% CI, 0.04–0.31) compared with NHW cases. On the other hand, AA cases were more likely to have H. pylori (OR, 2.33; 95% CI, 0.98–4.98) and chronic gastritis (OR, 3.29; 95% CI, 1.11–9.76) than NHW cases.
All Cases vs. all Controls
Cases were less likely than controls to be AA (unadjusted OR, 0.19; 95% CI, 0.13–0.28; Table 5). In univariate analysis of all cases and controls, cases were also more likely to be male and older, have a high WHR or hiatus hernia, and report GERD symptoms ≥10 years or PPI use. Cases were less likely to have H. pylori infection, active gastritis of the corpus, or chronic gastritis of the antrum and corpus than controls (Table 5). In the full multivariate model of all cases and controls, race, sex, H. pylori infection, hiatus hernia, active gastritis in the antrum, PPI use, and NSAID use were retained in the model (Table 5). Despite adjusting for risk factors, the effect of race as a predictor of BE was not attenuated compared with the unadjusted estimates, as the risk of BE was significantly lower in AAs compared with NHWs (adjusted OR 0.26; 95% CI, 0.17–0.39). Male sex, high WHR, hiatus hernia, and active gastritis of the antrum were associated with increased odds of BE. H. pylori infection was protective against the development of BE (adjusted OR 0.60; 95% CI, 0.38–0.97).
Source...