Use of Proton Pump Inhibitors and the Risk of Cholangitis

A Nationwide Cohort Study

Yang Won Min; Danbee Kang; Ju-Young Shin; Minwoong Kang; Joo Kyung Park; Kwang Hyuck Lee; Jong Kyun Lee; Kyu Taek Lee; Poong-Lyul Rhee; Jae J. Kim; Eliseo Guallar; Juhee Cho; Hyuk Lee


Aliment Pharmacol Ther. 2019;50(7):760-768. 

In This Article


During 4,212,003 person-years of follow-up, 58,863 (10.1%) participants had at least one PPI prescription. Compared to non-users, PPI users were older and more likely to have comorbidities at the start of follow-up (Table 1). PPI users were also more likely to have hepatobiliary diseases (1.5% vs 4.5%) and upper gastrointestinal diseases (34.7% vs 57.3%) and were to use immune-suppressants, antibiotics and systemic corticosteroids over follow-up than non-users (Table 2).

The number of incident cases of cholangitis observed over follow-up was 1,778 during unexposed periods (incidence rate 4.23 cases per 10 000 person-years) and 56 during PPI-exposure periods (incidence rate 44.37 cases per 10 000 person-years) (Table 2 and Figure 2). The age-, sex-, residential area- and income percentiles-adjusted HR for incident cholangitis comparing use vs no use of PPI was 6.06 (95% CI, 4.64-7.91). The association was essentially unchanged after adjustment for multiple confounders (fully adjusted HR 5.75; 95% CI, 4.39-7.54). When we conducted sensitivity analysis assuming that PPI exposure could only be responsible for cases of cholangitis occurring 14 days or more after treatment initiation, the HR for cholangitis associated with PPI use was 2.87 (95% CI, 1.77-4.63; Table S3). In addition, with the sensitivity analysis according to the duration of the period of residual effect, the results using 60 days (fully adjusted HR 5.59; 95% CI, 4.42-7.08) and 90 days (fully adjusted HR 5.07; 95% CI, 4.06-6.33) of residual effect period were similar to those using 30 days (Table S4). The positive association between PPI use and cholangitis was consistent in most pre-specified subgroups analysed (all P-values for interaction >.05; Figure 3). The association was stronger in participants who had less than five out-patient clinic visits during the year prior to the baseline health screening exam compared to participants who had more than five clinic visits (P value for interaction < .001).

Figure 2.

Cumulative incidence of cholangitis by the use of proton pump inhibitors (PPIs). Cumulative incidence was calculated using the Kaplan-Meier method. Participants who took PPIs contributed person-time to the exposed group while they were taking PPIs. Unexposed person-time was contributed by participants who did not take PPIs and participants who took PPI contributed unexposed person-time while they were not taking them (see text for details)

Figure 3.

Hazard ratios (95% confidence intervals) for incident cholangitis associated with the use of proton pump inhibitors (PPIs) in selected population subgroups. Adjusted for age (5-year categories), sex, residential area (metropolitan and rural), income level (≤30th, >30th-≤70th, >70th percentiles), BMI (<18.5, ≥18.5-<23, ≥23 - <25, ≥25 kg/m2 and unknown), smoking status (never, former, current and unknown) and frequency of alcohol intake (<1 time per week, 1–2 times per week, 3–4 times per week, almost every day and unknown) at the baseline health screening exam as time-fixed variables, for the number of out-patient clinic visits during the year prior to the baseline health screening exam as time-fixed variable, development of hepatobiliary diseases (including gallbladder stones, common/intrahepatic bile duct stones, biliary stricture/atresia, biliary cysts, liver cirrhosis), upper gastrointestinal diseases (gastro-oesophageal reflux disease, gastric ulcer and duodenal ulcer) and for the use of immune-suppressants, antibiotics and systemic corticosteroids over follow-up as time-varying variables.

Among participants with at least one PPI prescription, the excess risk for cholangitis was highest during the period of PPI use and decreased gradually thereafter (Table 3). Compared to the risk at >90 days after PPI discontinuation, the fully adjusted HR (95% CI) for cholangitis was 23.41 (95% CI, 12.32-44.46) during PPI use, 19.16 (95% CI, 12.43-29.52) within ≤30 days after PPI discontinuation and 15.83 (95% CI, 9.80-25.56) within >30 to ≤90 days after PPI discontinuation.