Preeclampsia and the Risk of Pancreatitis: A Nationwide, Population-Based Cohort Study

Background Preeclampsia is a multiple organ dysfunction during pregnancy, including hepatic, renal, and neurological dysfunction, and is defined as hypertension and proteinuria occurring after 20 weeks of pregnancy. Clinical features seen in preeclampsia are due to relatively poorly perfused placenta and maternal endothelial dysfunction. Some studies have found that preeclampsia may cause acute pancreatitis due to microvascular abnormalities and visceral ischemia. This retrospective cohort study used the Taiwanese National Health Insurance Research Databases (NHIRD) to study the relationship between preeclampsia and the risk of pancreatitis. Methods In total, 606,538 pregnant women were selected from the NHIRD between January 1, 1998 and December 31, 2010. They were divided into a preeclampsia cohort (n = 485,211) and a nonpreeclampsia cohort (n = 121,327). After adjusting for comorbidities that may induce pancreatitis, we analyzed and compared the incidence of pancreatitis in the two cohorts. Results The overall incidence of pancreatitis in the preeclampsia cohort was significantly higher than that in the control cohort (4.29 vs. 2.33 per 10,000 person-years). The adjusted HR of developing pancreatitis increased 1.68-fold (95% CI: 1.19-2.36) in the preeclampsia cohort. In addition, pregnant women with preeclampsia without comorbidities had a significantly high risk of pancreatitis (aHR = 1.83, 95% CI 1.27-2.63). The combined effect of preeclampsia and alcohol-related diseases resulted in the highest risk of pancreatitis (aHR = 43.4, 95% CI: 6.06-311.3). Conclusion Compared with patients without preeclampsia, the risk of pancreatitis in patients with preeclampsia is significantly increased after adjusting for demographics and comorbidities. The risk of pancreatitis is greatly increased when preeclampsia is accompanied by alcohol-related diseases, hepatitis C, gallstones, diabetes, or age of 26–35 years. Early identification and effective control of preeclampsia and the associated comorbidities can reduce the risk of pancreatitis and the associated morbidity and mortality.

The increased systemic inflammatory response plays a critical role in the pathogenesis of preeclampsia, leading to edema, extravasation, and increased damage to the vascular bed of the placenta, kidneys, and other organs [6]. This leads to poorly perfused placenta and maternal endothelial dysfunction [7]. These changes make preeclampsia a multiorgan syndrome dysfunction, with increased risks of various disorders, including chronic hypertension, diabetes mellitus, ischemic heart disease, cerebrovascular disease, kidney disease, thromboembolism, hypothyroidism, and even memory impairment [8]. Women with preeclampsia have an increased risk of life-threatening obstetric or medical complications. Globally, 10%-15% of maternal deaths due to pregnancy complications are directly related to preeclampsia/eclampsia [9].
Pancreatitis is the inflammation of the pancreatic glandular parenchyma, usually accompanied by abdominal pain and elevated serum pancreatic enzymes. Acute pancreatitis is the most common cause of hospitalization for gastrointestinal diseases in the United States [10]. Several conditions are related to acute pancreatitis, among which gallstones and chronic alcoholism account for approximately two thirds [11]. Systemic inflammatory response syndrome and organ failure are the main causes of death in the first two weeks of acute pancreatitis, and sepsis is the main cause of death after two weeks [12]. In a systematic review of studies of acute pancreatitis, the overall mortality rate is approximately 5% [13].
Severe preeclampsia can cause various systemic abnormalities, such as refractory abdominal pain, impaired liver function, severe hypertension, cerebral or visual disturbances, progressive renal insufficiency, and thrombocytopenia [14,15]. Some case reports have indicated a link between the development of pancreatic diseases and preeclampsia [16][17][18][19][20], but studies with a higher level of evidence examining this link are lacking. Therefore, this retrospective cohort study explored the relationship between preeclampsia and pancreatitis by using the Taiwanese National Health Insurance Research Database (NHIRD).

Materials and Methods
2.1. Data Source. This retrospective population-based study used data from the Taiwan NHIRD which covers medical benefit claims for over 23 million people (approximately 99% of Taiwan's population) [21]. The NHIRD contains the registry of beneficiaries and all medical service data. These claim files record the disease based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). For each beneficiary, a unique encrypted identification number is used to link all insurance information and health care records. This study was approved by the Ethics Review Board of China Medical University, Taichung (CMUH104-REC2-115).

Variables and Participants.
We identified all preeclampsia patients from the NHIRD corresponding to the International Classification of Disease, the Ninth Revision (ICD-9) codes 642.4-642.7 from 1998 to 2010. The date of preeclampsia diagnosis was assigned as the index date. For each woman with preeclampsia, we identified a pregnant woman without the history of preeclampsia for the comparison cohort, frequency matched by age (every 5 years span), and pregnancy year. Patients were excluded if they were <18 or >45 years of age and had a history of pancreatitis (ICD-9 code 577) before the index date. The confirmation of pancreatitis events was based on the database from 1998 to 2011 as the study endpoint. All study participants were followed up from the index date to the occurrence of endpoint, death, withdrawal from the database, or the end of 2011, whichever came first. We evaluated several comorbidities, which were defined as the history before the endpoint. They could be related to pancreatitis, namely, alcohol-related diseases (ICD-9

Statistical Analysis.
Distributions of age and comorbidities were compared between cohorts with and without preeclampsia and then tested using the chi-square test for categorical variables and Student's t -test for continuous variables. The cumulative incidence curve for pancreatitis was assessed using the Kaplan-Meier method, and intergroup differences were estimated using the log-rank test. We estimated the incidence densities of pancreatitis during followup in both cohorts. We evaluated the risk of pancreatitis for preeclampsia patients compared with the comparison cohort by using univariable and multivariable Cox proportional hazards models and presented by hazard ratios (HRs) and corresponding 95% confidence intervals (CIs). Multivariable models were adjusted for age and comorbidities. All statistical analyses were performed using SAS 9.4 (SAS Institute, Cary, NC, USA). We set the significant level at p < 0:05 for two-sided testing.

Results
The eligible study participants consisted of 17 263 patients in the preeclampsia cohort and 69 052 individuals in the nonpreeclampsia cohort. The baseline characteristics of all patients are summarized in Table 1. The mean age of patients with and without preeclampsia was 31:2 ± 5:19 and 31:0 ± 5:22 years, respectively, and 64.0% of the study participants were aged 26-35 years. Compared with patients without preeclampsia, those with preeclampsia were more likely to have biliary stones (1.74%), diabetes mellitus (5.02%), hyperlipidemia (1.81%), hypertension (3.85%), hepatitis B virus (0.74%), and hepatitis C virus (0.27%). The average followup duration was 7:03 ± 3:90 years for the preeclampsia cohort and 7:03 ± 3:92 years for the comparison cohort.

Discussion
This is the first nationwide, population-based study to investigate the incidence of pancreatitis after preeclampsia. Our results revealed that patients with a history of preeclampsia exhibited a 1.68-fold risk of pancreatitis than did those without preeclampsia, after adjusting for demographic characteristics and comorbidities.
The overall incidence of pancreatitis in the preeclampsia cohort was higher than that in the control cohort, even after adjusting for comorbidities associated with pancreatitis and even when only preeclampsia patients without comorbidities were considered (Table 2). Compared with the control cohort, the incidence of pancreatitis in patients with preeclampsia increased with age, especially in the 26-35 age group; this finding is consistent with studies indicating that advanced maternal age is prone to preeclampsia or pancreatitis [41,42].
The present study revealed that preeclampsia was an independent risk factor for pancreatitis. However, specific diseases, such as alcohol-related diseases, gallstones, or diabetes, were more influential than preeclampsia. Preeclampsia patients with a single specific comorbidity had a higher risk of pancreatitis than did nonpreeclampsia patients with that comorbidity (aHRs = 1:78 -1:89) ( Table 3), but the combined effects of preeclampsia and the specific comorbidity further increased the risk of pancreatitis. The risk was especially high for preeclampsia combined with alcohol-related diseases, as well as with hepatitis C, diabetes mellitus, and biliary stone.
The prevalence of pancreatitis in pregnancy is low and ranges between 1 in 1000 and 1 in 3000 deliveries [43], but maternal mortality with severe pancreatitis is high [44]. The incidence of fetal distress and fetal loss increases with the severity of pancreatitis [45]. Fortunately, early diagnosis of pancreatitis during pregnancy and improved maternal-infant intensive care has led to a declined in maternal and infant mortality rates [44]. The most common causes of acute pancreatitis in pregnancy are gallstones (67%-100% of pregnancy cases) [43], alcoholism, and hypertriglyceridemia [44]. Other causes include idiopathic pancreatitis, gestational hypertension, drug-induced pancreatitis, traumatic pancreatitis, and inherited diseases [42,44,46].

Gastroenterology Research and Practice
This study explored the relationship between preeclampsia and pancreatitis. Although no direct link was seen between preeclampsia and the pathology of pancreatitis, preeclampsia has been shown to cause global vascular endothelial dysfunction, which can lead to pancreatitis [7]. The pathophysiology of preeclampsia likely involves both maternal and placental factors, such as abnormalities in the uterine and placental circulations. The resultant ischemic placenta seems to introduce complex factors into the maternal circulation; this leads to maternal vascular endothelial dysfunction and eventually gives rise to the clinical manifestations of preeclampsia [4,[47][48][49][50]. Preeclampsia is associated with microvascular abnormalities, which may involve the cerebral, placental, hepatic, renal, and visceral circulation. Thus, the pancreatic vasculature may also get affected and cause acute pancreatitis, which leads to organized pancreatic necrosis [20]. Ramin et al. 43 investigated 9 cases of pancreatitis related to preeclampsia and found microthrombosis, intravascular coagulation, and vasculitis during preeclampsia, probably resulting in neurological, renal, hepatic, and placental diseases [43]. Another study reported that preeclampsia can cause acute edematous pancreatitis likely related to microvascular abnormalities and visceral ischemia [51]. Taken together, these findings indicate that maternal vascular endothelial dysfunction in patients with preeclampsia may increase pancreatic vascular system damage, eventually leading to pancreatitis. Further research is warranted to elucidate underlying biological mechanisms.
The difference in the cumulative incidence between patients with preeclampsia and control cohorts increased over time (Figure 1), indicating that the risk of pancreatitis in patients with a history of preeclampsia is long lasting. Because other comorbidities, such as alcoholism, gallstones, hypertension, dyslipidemia, diabetes, and overweight, further increase the risk of pancreatitis, both preeclampsia and comorbidities should be identified as soon as possible and treated effectively to reduce the patient's long-term morbidity and mortality.

Limitations
The strength of our research was its population-based design. However, the use of an observational database (i.e., NHIRD) has some inherent limitations.
First, the diagnostic accuracy was based on administrative data, making potential misjudging of preeclampsia and pancreatitis results inevitable. However, the Bureau of National Health Insurance randomly cross-checks medical records in all medical institutions to reduce error codes and misclassification bias. We identified hypertensive disorders during pregnancy and other comorbidities by using ICD-9-CM codes. In Taiwan, since the implementation of the National Health Insurance Program, prenatal care has been very well delivered [52]. The program includes 10 routine  antenatal checkups, which involve blood pressure measurement and urinary protein tests, allowing accurate and timely diagnosis of preeclampsia by obstetricians. Although we could not calculate the validity of diagnostic codes for hypertensive disorders in pregnancy, the high validity of the diagnostic codes of the NHIRD has been reported [53,54]. In addition, we used univariable and multivariable Cox proportional hazards models to assess the risk of pancreatitis and expressed them as HRs. Multivariable models were adjusted for age and comorbidities, including the exclusion of only hypertensive patients. The diagnosis of pancreatitis is typi-cally based on clear guidelines involving characteristic symptoms and signs, blood biochemistry results, and imaging findings [55] and is thus not prone to error. Second, the NHIRD does not include details of high alcohol consumption, body mass index, tobacco use, or other lifestyle-related factors that may be potential confounders. We tried to reduce these confounders by, for example, including alcohol-related diseases as a proxy of alcohol consumption and metabolic syndrome parameters (such as hyperlipidemia and diabetes) as an alternative indicator of body mass index. In general, the smoking rate among Rate † : incidence rate per 10000 person-years; adjusted HR † : multivariable analysis including age; * * p < 0:01, * * * p < 0:001.

Gastroenterology Research and Practice
Taiwanese women is low, being <2.5% among pregnant women [56]. Moreover, such a large population-based study may have neutralized this effect.

Conclusions
Compared with patients without preeclampsia, the risk of pancreatitis in patients with preeclampsia was found to be significantly high even after adjusting for certain demographic characteristics and comorbidities. Preeclampsia is an independent risk factor for pancreatitis, and the risk is further increased when accompanied by alcohol-related diseases, gallstones, diabetes, hepatitis C infection, or age of 26-35 years. Thus, timely identification and effective management of preeclampsia and the aforementioned comorbidities can reduce the risk of pancreatitis and the associated morbidity and mortality.