Table 2.
The association between PPI use and the risk of gastric cancer in the PCCIU database.
| Unadjusted | Adjusteda | Fully adjustedb | |||
|---|---|---|---|---|---|
| Case, n (%) | Control, n (%) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| PPI main analysis (removing 1 year before index) | |||||
| PPI user vs. non-user | 329/1119 (29.4) | 1213/5394 (22.5) | 1.45 (1.25, 1.68) | 1.48 (1.26, 1.73) | 1.49 (1.24, 1.80) |
| Male only | 167/639 (26.1) | 668/3082 (21.7) | 1.29 (1.05, 1.58) | 1.30 (1.05, 1.61) | 1.26 (0.97, 1.62) |
| Female only | 162/480 (33.8) | 545/2312 (23.6) | 1.66 (1.33, 2.06) | 1.74 (1.38, 2.19) | 1.84 (1.38, 2.47) |
| High-dose PPI user vs. non-user | 33/1119 (3.0) | 109/5394 (2.0) | 1.40 (0.94, 2.10) | 1.40 (0.93, 2.12) | 1.23 (0.76, 1.97) |
| Dose–response analysis (removing 1 year before index) | |||||
| 1–183 DDDs vs. non-user | 141/1119 (12.6) | 441/5394 (8.2) | 1.74 (1.41, 2.14) | 1.77 (1.43, 2.19) | 1.84 (1.43, 2.38) |
| 184–365 DDDs vs. non-user | 43/1119 (3.8) | 163/5394 (3.0) | 1.41 (1.00, 1.99) | 1.48 (1.04, 2.09) | 1.49 (0.97, 2.28) |
| 366–1095 DDDs vs. non-user | 81/1119 (7.2) | 352/5394 (6.5) | 1.21 (0.94, 1.56) | 1.23 (0.94, 1.59) | 1.20 (0.89, 1.64) |
| ≥1096 DDDs vs. non-user | 64/1119 (5.7) | 257/5394 (4.8) | 1.32 (0.98, 1.77) | 1.31 (0.97, 1.77) | 1.30 (0.91, 1.85) |
| P value for the trend | 0.003 | 0.004 | 0.026 | ||
| 1–6 prescriptions vs. non-user | 153/1119 (13.7) | 482/5394 (13.7) | 1.71 (1.40, 2.10) | 1.74 (1.42, 2.14) | 1.85 (1.44, 2.37) |
| 6–12 prescriptions vs. non-user | 44/1119 (3.9) | 167/5394 (3.1) | 1.42 (1.01, 2.00) | 1.46 (1.04, 2.07) | 1.39 (0.91, 2.13) |
| 12–36 prescriptions vs. non-user | 90/1119 (8.0) | 366/5394 (6.8) | 1.31 (1.02, 1.68) | 1.33 (1.04, 1.72) | 1.35 (1.00, 1.82) |
| ≥36 prescriptions vs. non-user | 42/1119 (3.7) | 198/5394 (3.7) | 1.07 (0.75, 1.53) | 1.06 (0.74, 1.52) | 0.97 (0.64, 1.47) |
| P value for the trend | 0.008 | 0.010 | 0.073 | ||
| PPI user vs. non-user when removing prescriptions in specific duration | |||||
| Removing 2 years before index | 212/862 (24.6) | 878/4126 (21.3) | 1.20 (1.00, 1.43) | 1.19 (0.99, 1.43) | 1.13 (0.91, 1.40) |
| Removing 3 years before index | 137/644 (21.3) | 620/3062 (20.3) | 1.04 (0.84, 1.28) | 1.04 (0.83, 1.29) | 1.03 (0.80, 1.33) |
| Removing 4 years before index | 87/474 (18.3) | 421/2235 (18.8) | 0.93 (0.71, 1.21) | 0.92 (0.70, 1.21) | 0.89 (0.65, 1.22) |
| Removing 5 years before index | 55/345 (15.9) | 267/1611 (16.6) | 0.90 (0.65, 1.25) | 0.90 (0.64, 1.26) | 0.82 (0.55, 1.22) |
| PPI user in specific time periods before index date/cancer diagnosis date | |||||
| 0–1 y before indexc | 664/1119 (59.3) | 1088/5394 (20.2) | 6.32 (5.44, 7.33) | 6.79 (5.82, 7.91) | 7.04 (5.75, 8.61) |
| 1–2 y before indexd | 259/1119 (23.2) | 926/5394 (17.2) | 1.45 (1.23, 1.70) | 1.47 (1.25, 1.74) | 1.51 (1.23, 1.84) |
| 2–3 y before indexe | 165/862 (19.1) | 670/4126 (16.2) | 1.20 (0.99, 1.46) | 1.20 (0.98, 1.47) | 1.15 (0.90, 1.46) |
| 3–4 y before indexf | 107/644 (16.6) | 481/3062 (15.7) | 1.04 (0.82, 1.31) | 1.02 (0.81, 1.30) | 1.00 (0.76, 1.33) |
| 4–5 y before indexg | 68/474 (14.4) | 329/2235 (14.7) | 0.93 (0.70, 1.24) | 0.92 (0.69, 1.23) | 0.96 (0.68, 1.35) |
| PPI new user vs. PPI non-new userh | 376/1119 (33.6) | 235/5394 (4.4) | 10.93 (9.01,13.25) | 11.11 (9.14, 13.51) | 10.98 (8.47, 14.23) |
| Omeprazole user vs. non-user | |||||
| Removing 1 year before the index | 201/1119 (18.0) | 774/5394 (14.4) | 1.30 (1.09, 1.55) | 1.29 (1.07, 1.54) | 1.21 (0.97, 1.50) |
| Removing 2 years before the index | 123/862 (14.3) | 548/4126 (13.3) | 1.06 (0.85, 1.32) | 1.03 (0.82, 1.29) | 0.92 (0.70, 1.20) |
| Removing 3 years before the index | 78/644 (12.1) | 373/3062 (12.2) | 0.95 (0.73, 1.24) | 0.95 (0.72, 1.25) | 0.89 (0.65, 1.23) |
| Lansoprazole user vs. non-user | |||||
| Removing 1 year before the index | 169/1119 (15.1) | 610/5394 (11.3) | 1.40 (1.16, 1.69) | 1.42 (1.17, 1.73) | 1.49 (1.18, 1.88) |
| Removing 2 years before the index | 114/862 (13.2) | 447/4126 (10.8) | 1.24 (0.99, 1.56) | 1.24 (0.99, 1.57) | 1.27 (0.97, 1.66) |
| Removing 3 years before the index | 75/644 (11.7) | 319/3062 (10.4) | 1.11 (0.85, 1.46) | 1.12 (0.84, 1.48) | 1.17 (0.85, 1.62) |
| Sensitivity analysis (PPI user vs. non-user) | |||||
| Additionally adjusted for H2RAi | 329/1119 (29.4) | 1213/5394 (22.5) | 1.45 (1.25, 1.68) | 1.41 (1.20, 1.65) | 1.43 (1.18, 1.73) |
| PPI user vs. H2RA userj | 329/431 (76.3) | 1213/1604 (75.6) | 1.18 (0.85, 1.65) | 1.24 (0.88, 1.75) | 1.34 (0.85, 2.09) |
| Adjusted for lifestyle using multiple imputationk | 329/1119 (29.4) | 1213/5394 (22.5) | 1.45 (1.25, 1.68) | 1.48 (1.26, 1.73) | 1.47 (1.26, 1.72) |
| Additionally not adjusted for peptic ulcer and oesophagitisl | 329/1119 (29.4) | 1213/5394 (22.5) | 1.45 (1.25, 1.68) | 1.45 (1.25, 1.69) | 1.44 (1.20, 1.74) |
aStudy matched on age, sex and general practice, and the model contains obesity, comorbidities in the exposure period (including diabetes, coronary heart disease, myocardial infarction, heart failure, peripheral vascular disease, cerebrovascular disease, cerebrovascular accident, chronic obstructive pulmonary disease, mental illness, liver disease, peptic ulcer and oesophagitis) and other medication uses in the exposure period (statins and aspirin).
bAdditionally adjusted for alcohol and smoking.
cMedication use in the year prior to diagnosis/index date restricted to individuals with at least 3 years of records.
dMedication use in the year from 2 to 1 year prior to diagnosis/index date restricted to individuals with at least 3 years of records.
eMedication use in the year from 3 to 2 years prior to diagnosis/index date restricted to individuals with at least 4 years of records.
fMedication use in the year from 4 to 3 years prior to diagnosis/index date restricted to individuals with at least 5 years of records.
gMedication use in the year from 5 to 4 years prior to diagnosis/index date restricted to individuals with at least 6 years of records.
hProportion of cases and controls who used PPI in the year before diagnosis and who had not previously used PPI.
iAdditionally adjusted for H2RA.
jUsing only H2RA users as an active comparator.
kUsing multiple imputation to adjust for alcohol and smoking.
lRemoving the peptic ulcer and oesophagitis adjustment from the main model.