Multiple substance use and the risk of pancreatitis: a systematic review
Department of Biomedical Sciences, Cedars-Sinai Medical Center, 6500 Wilshire Blvd, 15th Floor, 1544B, Los Angeles, CA 90048, USA
Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Division of Gastroenterology, Hepatology and Nutrition, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA
Department of Biomedical Sciences, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA
Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Department of Biomedical Sciences, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Abstract
Background:
The impact of multiple substance use on the risk of pancreatitis remains underexplored.
Objective:
To systematically review peer-reviewed observational studies assessing the association of multiple substance use with the risk of acute pancreatitis (AP) or chronic pancreatitis (CP) in adults.
Design:
We conducted a systematic review informed by the Preferred Reporting Items for Systematic Review and Meta-Analyses guideline.
Data sources and methods:
EMBASE, MEDLINE, and PsycINFO were searched up to March 2024. Reference lists of included studies were reviewed. From 5205 records identified, 181 relevant records were evaluated in full text. Studies evaluating the association of ⩾2 substances, including tobacco, alcohol, cannabis, and illicit substances, with AP or CP were included. Data were extracted by one reviewer, with quality control by a second reviewer. Quality assessments were independently conducted by two reviewers, with differences resolved by a third.
Results:
Of 11 included studies, 6 investigated AP as the outcome and 5 examined CP. Among AP studies, 5 comparing smoking and alcohol to alcohol-only use showed high heterogeneity (I2 = 90.9%), with relative risks (RRs) from 1.40 to 11.40. One study examining cannabis and alcohol versus alcohol found a lower risk of AP in cannabis users. Among CP studies, four comparing smoking and alcohol to alcohol-only use were heterogeneous (I2 = 81%) with odds ratios 1.21–31.50. Where examined, smoking increases the risk of AP and CP in a dose-dependent fashion. Heavy alcohol users demonstrated a significant increase in CP risk across all smoking categories in one study.
Conclusion:
Combined alcohol and tobacco use increases pancreatitis risk compared to single substance use, despite heterogeneity in RRs and exposure definitions. Evidence suggests a dose-dependent impact of smoking on pancreatitis risk when added to alcohol. Studies on the impact of a combination of other substance use on pancreatitis risk are needed.
Trial registration (PROSPERO):
CRD42024503677.
Untitled section
Keywords: adult, cannabis, ethanol, pancreatitis, risk, smoking, substance-related disorders, tobacco use
Article notes
Untitled section
Received 2025 Jan 28; Accepted 2025 Jul 21; Collection date 2025.
Introduction
Pancreatitis is inflammation of the pancreas that can present as acute pancreatitis (AP), recurrent acute pancreatitis (RAP), or chronic pancreatitis (CP). After an initial AP episode, 22% of patients develop RAP, and 36% of RAP patients develop CP, compared to 10% who progress from a first-AP episode to CP. 1 Globally, the incidence of AP is estimated at 34 cases per 100,000 person-years and 10 cases per 100,000 person-years for CP. 2 Most patients with any form of pancreatitis require hospitalization, which places a significant burden on healthcare systems.3,4 High hospitalization rates in the United States are primarily driven by AP cases, with over 280,000 annual visits. 5 Despite advances in treatment and prevention, pancreatitis remains a major public health problem, 6 with risks of organ failure, pancreatic cancer, and mortality. 7
Existing systematic studies mainly examine single-substance use, such as tobacco-only or alcohol-only use, with findings of increased risk of pancreatitis with either substance.3,8 Multiple substance use, which is the use of more than one substance either simultaneously or within a brief timeframe, 9 has been overlooked, despite its potential synergistic impact on pancreatitis risk. Some epidemiological studies on multiple substance use and pancreatitis reveal surprising associations. For example, one study found that concurrent cannabis and heavy alcohol use were associated with a lower incidence of AP and CP (adjusted odds ratio (aOR): 0.50; 95% confidence interval (CI): 0.48–0.53, p < 0.0001) and (aOR: 0.77; 95% CI: 0.71–0.84, p < 0.0001), respectively. 10 Conversely, current smokers consuming ⩾400 g of alcohol per month had a twofold increase in pancreatitis risk (relative risk (RR): 2.10; 95% CI: 1.38–3.19, p < 0.01) compared to never smokers who consumed the same level of alcohol. 11 Alcohol, tobacco, cannabis, and illicit drugs, such as amphetamines and cocaine, are among the most widely used substances globally. 12 Multiple substance use is associated with higher substance dependence, lower cessation rates, elevated cancer risk, and mental health concerns,12,13 which may lead to worse health outcomes for pancreatitis patients compared to single-substance use. 14
Therefore, this study aims to systematically review existing literature and synthesize evidence on the association between multiple substance use and the risk of pancreatitis. By examining multiple substance use, this research could inform interventions directed at addressing cessation challenges and disease exacerbation in pancreatitis patients.
Methods
Eligibility criteria
Our inclusion criteria consisted of (a) studies on multiple substance use, especially those that report on the most common patterns of multi-substance use among adults globally, including alcohol, tobacco, cannabis, and illicit drugs, (b) studies that report on the risk or association between multiple substance use and pancreatitis, (c) studies examining outcomes of AP or CP, (d) studies focused on adults (⩾18 years), (e) peer-reviewed articles quantitatively assessing the association of multiple substance use and AP or CP risk, and (f) articles published in English. We excluded studies that only examined the impact of one substance or did not examine the joint effects of multiple substances. Case reports and meta-analyses were excluded.
Search strategy and information sources
Following registration in PROSPERO (ID: CRD42024503677), a preliminary search on PubMed, including MEDLINE, was conducted to identify relevant search terms. These terms then guided our systematic searches across three databases: EMBASE, MEDLINE, and PsycINFO. The search strategies for the three databases, spanning from the beginning of indexing to the date of last search (March 2024), are detailed in Supplemental Table 1. Briefly, we combined the Medical Subject Heading terms of pancreatitis, epidemiologic studies, and substance use, with “alcohol,” “smoking,” “tobacco,” “cannabis,” “marijuana,” and “illicit” added as key search terms for identification in abstracts and titles. The reference lists of included publications were also examined to identify additional eligible articles. This review adhered to Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) reporting guidelines (Supplemental Table 2). 15
Selection process
Our title and abstract review process employed an artificial intelligence (AI) tool for screening. 16 Following a literature search, all references were imported into Rayyan software to remove duplicates and then into ENDNOTE to deduplicate to avoid skewing AI-based classifications. References were then imported into the AI tool “ASReview” (version 1.6) for screening. ASReview was set to default mode to classify the relevance of articles, using Term Frequency-Inverse Document Frequency (TF-IDF) for feature extraction, Naïve Bayes (NB) as classifier, query strategy as maximum, and balance strategy as dynamic resampling (double), providing a faster and excellent performance. Combining NB and TF-IDF yields superior performance than other configurations. 16 We trained the AI algorithm with four manually selected relevant studies and four AI-randomly selected irrelevant studies. Once trained, ASReview generated an initial ranking of all unlabeled articles (i.e., articles pending eligibility decisions) based on the highest to lowest likelihood of relevance. A reviewer (Y.J.) then assessed the title and abstract of the highest-ranking article in the unreviewed stack and marked it as “relevant” or “irrelevant” according to our eligibility criteria. The AI incorporated this feedback to adjust rankings and presented the next highest-ranked article for review. The interactive process by which AI makes rankings and then the reviewer makes rankings is termed “researcher-in-the-loop.” 16 The process continued until we reached a predefined stopping criterion of 245 consecutive irrelevant articles.
Articles labeled as relevant during the title and abstract screening were then reviewed at the full-text stage by two independent reviewers (Y.J. and E.A.A.). Any disagreements about the inclusion of studies were resolved by consensus between the two reviewers.
Data extraction
The PECO (Patient, Exposure, Comparison, and Outcome) framework was used to capture key elements that align with our research aim. 17 The following key variables were extracted from the articles: (a) first author’s name, (b) year of publication, (c) country where the study was conducted, (d) study aim, (e) study design, (f) study period (recruitment/follow-up), (g) sample size for each comparison group and the number of outcomes in each group, (h) substance definition and category boundaries, (i) outcome measures including RR estimates, measures of association, 95% CI, and pancreatitis definition (AP or CP), (j) adjustment factors and/or matching factors, (k) modeling methods, (l) comparison group, (m) subgroup analysis, and (n) interaction analysis. Data were extracted into REDCap by EAA, and accuracy was checked by Y.J. To obtain additional data, we contacted the lead author of the publication.
Overlapping definitions
Quality assessment
The quality of studies was assessed using the Joanna Briggs Institute (JBI) critical appraisal tools according to study design. Based on prior literature, 20 we ranked studies as follows: (1) high quality: a score of ⩾7 on a 10-point scale or ⩾70% if the maximum score is <10; (2) moderate quality: scores between 4 and < 7 (or ⩾ 40% and < 70%); and (3) low quality: a score of <4 or below 40%. Each question was assigned scores on a binary scale: “1” for yes or “0” for no, unclear, or not applicable. 21 Two reviewers (Y.J. and E.A.A.) independently conducted quality assessment, and discrepancies were resolved by a third reviewer (C.Y.J.).
Statistics
Statistical analyses were performed using R software, version 4.2.1, 22 with statistical significance set at p < 0.05. Heterogeneity across studies was assessed using I2 statistics, where I2 values ranged from 0% to 100%, with 25% indicating low heterogeneity, 50% moderate, and 75% high. 23 Due to high heterogeneity and variable exposure and reference categories, we did not compute a summary estimate. Funnel plots were visually inspected for publication bias. Given the limited sample size and heterogeneity by exposure definitions, Egger’s and Begg’s tests to detect publication bias were not conducted. Forest plots were created by grouping studies into clusters based on comparison categories. Studies with only one study in the comparison category were excluded from the forest plot. In two studies,19,24 we computed the RRs from the contingency tables. Furthermore, a continuity correction was applied, adding 0.5 to estimate the risk in Lowenfel’s study, 19 which had a group with a 0 cell count. Sensitivity analyses were not performed as the variable exposure and reference group definitions made it infeasible to attribute outliers.
Results
Study selection and characteristics
Figure 1 presents the PRISMA flow diagram for our search results and selection process. A total of 6811 records were identified, including 4816 from EMBASE, 1910 from MEDLINE, and 55 from PsycINFO. After removing 1606 duplicates, 5205 records were imported into the AI software. Of these, 181 were classified as relevant through AI-based title and abstract screening and confirmation by Y.J., followed by full-text review by two independent reviewers. Ultimately, 11 articles were included, with 1 by Haber et al., 18 and sourced from the reference list. Although another study met the inclusion criteria, 25 it was excluded due to a lack of detail on multiple substance use.
The characteristics of the 11 included studies are summarized in Table 1, such as study design, study population, exposure and outcome definitions, and comparison group. Additional characteristics such as analytic sample size and adjustment factors are shown in Supplemental Table 3. The study designs consisted of one cross-sectional study, five case–control studies, and five cohort studies. Of these, six studies examined AP as the outcome,10,11,24,26–28 and five studies examined CP as the outcome,10,18,19,29,30 of which two included RAP cases.18,19 One study assessed both AP and CP as outcomes. 10 Another study assessed combined AP, RAP, and CP as outcomes. 31 The studies compared the following multiple substance combinations and reference groups: cannabis and alcohol versus alcohol, 10 smoking and alcohol versus smoking, 24 smoking and alcohol versus alcohol,10,11,18,19,24,27,30 and smoking and alcohol versus neither.26,28 No studies examined other substance combinations or comparisons.
| First author, pub year | Study design | Study population | Exposure definition | Comparison group | Outcome definition |
|---|---|---|---|---|---|
| Adejumo, 2019 | Cross-sectional | 18,086,789 adults of the Nationwide Inpatient Sample dataset from 2012 to 2014. | (i) Cannabis: ICD-9-CM: 305.2×, 304.3× (ii) Alcohol: ICD-9-CM: 303.×, 305.0× (iii) Tobacco: ICD-9-CM: 305.1, 989.84, V15.82. | Alcohol users and non-cannabis users. | AP and CP (ICD-9-CM: 577.0 and 577.1, respectively). |
| Ramzan, 2022 | Case–control | 294 participants of a research conducted in Mayo Hospital, Lahore, Pakistan for 6 months (November 25, 2019 to May 24, 2020). | (i) History of smoking (yes or no) (ii) Alcohol use (yes or no) | (i) Smoking and non-alcoholism (ii) Non-smoking and alcoholism | Mention of the use of an operational definition for AP with no further detail provided. |
| Sadr-Azodi, 2012 a | Prospective cohort | 84,667 participants (female SMC and COSM). Recruitment for SMC (1987–1990, and 1997) and COSM (1997). SMC and COSM followed from 1998 to 2009. | Tobacco and alcohol: (i) Current, former, and never smokers, (ii) Current <20 py, current ⩾20 py, former <20 py, former ⩾20 py, and never-smokers. Alcohol drinkers: <400 or ⩾400 g/month | Never smokers + alcohol drinkers | Outcome was a first-time event of AP after January 1, 1998, with no prior history of CP: (i) AP (ICD10: K850e9, ICD-9: 577A, ICD-8: 577.00e08) (ii) CP (ICD-10: K86e9, ICD-9: 577B, ICD-8: 577.09e19) |
| Yadav, 2009 | Prospective Case–control | 1695 participants of the North American Pancreatitis Study 2 from 20 secondary or tertiary pancreatic care centers across the United States between 2000 and 2006. | Tobacco and alcohol, including: (i) smoking amount as <1 or ⩾1 PPDand quantified as pys (<12, 12–35, >35), (ii) Smoking status as “never” (smoked <100 cigarettes in lifetime) or “ever” (smoked >100 cigarettes). Ever smokers as “past” or “current” smokers, (iii) light drinkers (⩽0.5 drinks/day or ⩽3 drinks/week), (iv) moderate drinkers (>0.5–1 drinks/day or 4–7 drinks/week for females; >0.5–2 drinks/day or 4–14 drinks/week for males), (v) heavy drinkers (>1–<5 drinks/day or 8–34 drinks/week for females; >2–<5 drinks/day or 15–34 drinks/week for males), and (vi) very heavy drinkers (⩾5 drinks/day or ⩾35 drinks/week for both genders). | (i) Never smokers + light alcohol (ii) Never smokers + moderate alcohol (iii) Never smokers + heavy/very heavy alcohol | (i) RAP was defined by the presence of ⩾2 episodes of AP, (ii) CP by the presence of changes primarily on imaging studies or histology. |
| Setiawan, 2016 | Prospective cohort | 145,886 participants of a multiethnic cohort of African Americans, Native Hawaiians, Japanese Americans, Latinos, and Whites enrolled at baseline between 1993 and 1996. The follow-up period was from 1993 to 2012. | (i) Smoking as never, past, and current. Past and current were stratified by <20 and ⩾20 pack-years, (ii) Alcohol use frequency obtained in the year prior to baseline administration. Alcoholic drinks as regular beer, light beer, red wine/pink wine (including champagne and sake), and hard liquor. Nine intake categories spanning from “never” to “4 or more times per day” and information on usual serving size was also requested. Total alcohol intake is calculated by multiplying the volume of a drink by the percentage of alcohol. For males, the total alcohol intake is categorized as non-drinkers, <24, 24–⩽48, and >48 g ethanol/day, and for females as non-drinkers, <12, 12–<24, and ⩾24 g ethanol/day. | (i) Never smokers + non-drinkers | The outcome is non-GS pancreatitis (combination of non-GS AP, and RAP/CP): (i) ICD, 9th Revision, code 577.0 for AP or 577.1 for CP to determine pancreatitis cases, (ii) Subsequently cases were categorized as AP if they had one AP hospitalization and RAP if they had >1 AP hospitalization, (iii) AP group was further categorized into gallstone and non-gallstone-related types using ICD-9 codes for gallstone (574.×) from the same pancreatitis hospitalization claim, as well as procedure codes for cholecystectomy (ICD-9: 51.2× and Current Procedural Terminology codes: 47480, 47490, 47562, 47563, 47564, 47600, 47605, 47610, 47612, 47620, 56340, 56341, 56342), (iv) For each pancreatitis case, the event date was defined as the date of the first hospitalization in which pancreatitis was the principal diagnosis |
| Morton, 2004 | Prospective cohort | 128,934 adults of the Kaiser Permanente Medical Care Program in Northern California were recruited from 1978 to 1985. Participants were followed until the first of December 31, 1998, death or other health plan termination, or first pancreatitis hospitalization. | Tobacco and alcohol, including: (i) Cigarette smoking categorized as never smoked, and ex-smokers (<1 and ⩾1 PPD). (ii) Alcohol use in the past year as drinks/day included two categories (a) Never drank alcohol and (b) Ex-drinker (<Once per month, >Once per month but <1 drink/day, 1–2 drinks/day, and ⩾3 drinks/day). | (i) Never smokers and alcohol group with AP | Outcome is alcohol AP group: Defined by a primary or secondary diagnosis of pancreatitis (Code 577). Excluded people with a prior pancreatitis at baseline to focus on first-time cases. Pancreatitis was confirmed if history of symptoms and physical examination were compatible with elevated serum amylase. In cases without elevated serum amylase, evidence of pancreatitis by imaging or direct examination of the organ was required. Diagnoses and etiological categories were verified by trained medical record analysts and physician investigators. |
| Lin, 2000 | Case–control | Cases were male patients aged 26–79 years and newly diagnosed with CP in 11 university hospitals and their 52 affiliated hospitals throughout Japan. Controls were selected from in/outpatients with other diseases. The study was conducted from July 1997 to December 1998. | (i) Smoking status categorized as: non-smokers, ex-smokers, and current smokers. Among current smokers, smoking intensity was divided based on the average daily cigarette use (<20, 20–39, and ⩾40) (ii) Drinking status categorized as: non-drinkers, ex-drinkers, and current drinkers. For current drinkers, ethanol intake (g/day) was calculated for each beverage based on (reported daily frequency) × (amount of the beverage in Japanese drink) × (ethanol in one Japanese drink (29 g)). Thus, categorized as ⩾29 and 0–28 g/day. | Smoked (0–19 cigarettes/day) + alcohol (0–28 g/day) | Adopted the diagnostic criteria proposed by the Japan Pancreas Society in 1995 to define CP. |
| Lindkvist, 2008 | Prospective cohort | 33,346 individuals who participated in the Malmö Preventive Project, with a recruitment period from 1974 to 1992. Participants were followed until December 31, 1999. Participants were followed until they had an AP attack or died. | (i) Smoking status categories were never smokers, former smokers, and current smokers, (ii) Mm-MAST category involved a scoring system, whereby a “yes” was 1 point and a “no” was no points, except for the question “Are you a teetotaler?” where the scoring was reversed. Alcohol consumption was classified as “low risk” for those with a score of 0–1, “intermediate or medium risk” for those with a score of 2–3 and “high risk” for those with a score of 4 or more. | Never smokers with low-risk alcohol use | The diagnosis of AP was accepted in cases with a history of acute abdominal pain and elevated serum amylase, or when evidence of the disease was found during laparotomy or autopsy. Patients with a prior AP attack or CP history were excluded. |
| Lowenfels, 1987 | Retrospective case–control | A total of 181 participants of a study that spanned from 1980 to 1984 across hospitals that serve about half of Arizona’s Native American Population. | (i) Current smoking history status as non-smokers, smokers of <1 PPD, and smokers of 1 or more PPD, (ii) Alcohol use as: non-drinkers, moderate drinkers, and heavy drinkers. | Non-smokers with cirrhosis who consume alcohol | Outcome is RAP cases with acute on CP related to alcohol use. Patients presented with abdominal pain, nausea, and vomiting, along with at least one elevated amylase. Alcoholic pancreatitis patients had a history of alcohol abuse, previous attacks of pancreatitis, and/or presence of pancreatic calculi. |
| Munigala, 2015 | Retrospective cohort | 484,624 patients of a cohort of VHA patients identified between 1998 and 2007. Patients were followed from 2000 to 2007. | (i) Smoking status was defined as the presence of ICD-9 codes 305.1 or V15.82, (ii) Alcohol abuse/dependence included ICD-9 codes 303, 303.0, 303.9, 305.0. | (i) Alcohol group and non-tobacco users (ii) Alcohol group and non-tobacco users | The outcome was AP, identified by ICD-9 code 577.0. Also, two or more AP diagnoses coded within a 15-day timeframe were regarded as a single episode. |
| Haber, 1993 | Case–control | 99 participants comprising inpatients and outpatients of public hospital facilities in the same Sydney area. The study period was not mentioned. | (i) Daily cigarette consumption was dichotomized at 20 cigarettes/day, (ii) lifetime consumption of tobacco: 1 py corresponds to 20 cigarettes/day for 1 year, (iii) smoking status including current, former, and never smokers, (iv) the measure of cumulative alcohol consumption was derived from the product of daily alcohol intake and the number of years of heavy drinking. | Alcoholics without pancreatitis | The outcome is alcoholics with CP, defined as the presence of at least one of the following criteria: (i) previous acute attacks of pancreatitis with severe abdominal pain, abdominal tenderness, and serum amylase or lipase levels exceeding three times the upper limit of normal (ii) chronic abdominal pain with pancreatic calcification visible on abdominal X-ray or CT scan (iii) severe abdominal pain and endoscopic pancreatogram changes in moderate or severe pancreatitis based on Sarner and Cotton’s criteria (iv) histopathological changes of CP in tissue obtained at operation in a patient with a history of severe abdominal pain, suggesting pancreatitis. (v) pancreatitis related to alcohol abuse |
Exposure definitions for smoking and alcohol varied across studies. Smoking status was grouped into current, former, and never smokers,11,18,26,31 with Morton et al. 27 reporting only former and never use. Smoking was also categorized using pack-years, with thresholds such as <12, 12–35, and >35 pack-years, 30 or <20 and ⩾20 pack-years.11,31 Daily smoking was categorized as <20 or ⩾20 cigarettes/day by Lin et al. 29 while Lowenfels et al. 19 used <1 or ⩾1 pack/day. Alcohol use was quantified in grams/day by Setiawan et al. 31 as <24, 24–⩽48, and >48 for men, with lower thresholds (<12, 12–<24, and ⩾24) for women. Yadav et al. 30 classified alcohol use as light, moderate, and heavy/very heavy drinking, and Lindkvist et al. 26 reported alcohol use by risk levels (low, medium, and high). Dichotomous alcohol use categories were reported by Sadr-Azodi et al. 11 (<400 and ⩾400 g/month) and Morton et al. 27 (⩾29 and 0–28 g/day). In addition, two studies relied on the International Classification of Diseases-9 (ICD-9) codes for identifying smoking and alcohol use.10,28 Ramzan et al. 24 used binary categories (yes or no) to indicate smoking and alcohol use (Table 1 and Figures 2 and 3).
Acute pancreatitis
Figure 2 presents a forest plot of RRs of studies on multiple substance use and AP risk. Among five studies,10,11,24,27,28 assessing smoking and alcohol use to alcohol-only use, high heterogeneity (I2 = 90.9%, p < 0.0001) was observed, and the RRs ranged from 1.40 to 11.40. Two studies reported on current and former smoking combined with alcohol use compared to exclusive alcohol use.11,27 Specifically, former smokers who smoked <20 pack-years and drank alcohol had a RR of 1.97 (95% CI: 0.91–4.25) as compared to alcohol-only users, which increased to 3.96 (95% CI: 1.87–8.39) for those with ⩾20 pack-years of smoking who drank alcohol. 11 Also, former smokers who smoked <1 pack/day and drank alcohol had a RR of 5.80 (95% CI: 3.23–10.41), which increased to 11.40 (95% CI: 6.47–20.10) for those who smoked ⩾1 pack/day and drank alcohol. 27 Compared to never smokers who consumed alcohol, current smokers who smoked <20 pack-years and drank alcohol faced a 2.13-fold risk of AP (95% CI: 0.84–5.40), which increased to 4.12 (95% CI: 1.98–8.59) for those with ⩾20 pack-years of smoking who drank alcohol. 11 For comparisons of smoking and alcohol use to neither, two studies reported high heterogeneity (I2 = 91.4%, p < 0.0001) with RRs of 1.08 and 7.45.26,28 In the study reported by Lindkvist et al., 26 former smokers with medium-risk alcohol use showed RR of 1.08 (95% CI: 0.50–2.33) compared to non-smokers with low risk of alcohol use, while former smokers with high-risk alcohol use showed a RR of 3.05 (95% CI: 0.92–10.14), though these findings were not significant. Current smokers with medium risk for alcohol abuse showed a 2.48 times higher risk (95% CI: 1.52–4.05) of AP, which increased to 4.85 (95% CI: 2.66–8.83) for those with high-risk alcohol use compared to the non-smoking group with low risk of alcohol use. 26 Detailed findings from additional studies not presented in the forest plot are summarized in Supplemental Table 4. We identified one study 24 reporting that combined alcohol and tobacco use, compared to tobacco use alone, was associated with a RR of 2.51 (95% CI: 0.65–9.68). We also found one study comparing cannabis use among persons with alcohol use, 10 reporting a reduced risk (RR: 0.50; 95% CI: 0.48–0.53) of AP attributable to concurrent cannabis and alcohol use compared to alcohol-only use.
Chronic pancreatitis
Figure 3 presents a forest plot of studies examining multiple substance use and CP risk. In studies examining combined smoking and alcohol use versus alcohol use alone,10,18,19,30 heterogeneity was high (I2 = 81%, p < 0.0001) with ORs ranging from 0.52 to 31.50. Yadav et al. 30 highlighted varying ORs based on smoking intensity and alcohol consumption. Among moderate alcohol drinkers, the ORs for CP were 1.87 (95% CI: 0.82–4.27) for smokers with <12 pack-years and 2.17 (95% CI: 0.96–4.90) for 12–35 pack-years, and an OR of 7.59 (95% CI: 2.93–19.65) for smokers with >35 pack-years compared to non-smokers. 30 For heavy or very heavy alcohol drinkers, a significantly increased risk of CP was observed across all smoking categories including <12 pack-years (OR: 3.13; 95% CI: 1.25–7.86), 12–35 pack-years (OR: 4.47; 95% CI: 1.93–10.35), and >35 pack-years (OR: 13.41; 95% CI: 5.23–34.39) compared to never smoking, suggesting a dose-dependent risk of CP. 30 Lowenfels et al. 19 presented the number of male alcoholic CP cases and alcoholic liver cirrhosis controls by smoking exposure categories, with which we estimated ORs of 1.30 (95% CI: 0.32–5.32) for those who smoked <1 pack/day with alcohol-related pancreatitis and 31.5 (95% CI: 7.7–575) for those who smoked ⩾1 pack/day compared to non-smokers. Supplemental Table 5 additionally summarizes the risk estimates for CP by substance use categories among studies not represented in the forest plot. Compared to alcohol-only users, former smokers who consumed alcohol showed an OR of 1.33 (95% CI: 0.15–11.89), while current smokers who consumed alcohol showed an OR of 1.21 (95% CI: 0.26–5.74). 18 One study examined smoking and alcohol use versus neither. 29 Those that smoked (<20 cigarettes/day) and drank ⩾29 g/day reported an OR of 4.70 (95% CI: 1.52–14.51), while the OR was higher (OR: 7.30; 95% CI: 2.42–22.05) for those who smoked ⩾20 cigarettes/day and drank alcohol of ⩾29 g/day. Another study found that concurrent cannabis and alcohol use was significantly associated with a lower risk of CP (RR: 0.77; 95% CI: 0.71–0.84) compared to alcohol use alone. 10
Combined pancreatitis
Only one study assessed the combined risk of pancreatitis (AP, RAP, and CP) associated with smoking and alcohol use compared to neither (Supplemental Table 6). 31 Among male past smokers, neither daily consumption of ⩽48 g of alcohol nor those with consumption of >48 g of alcohol daily showed a significantly increased risk of pancreatitis. By contrast, current male smokers consuming ⩽48 g of alcohol daily showed a significantly increased hazard ratio (HR) of 1.54 (95% CI: 1.13–2.09), which further increased to 2.06 (95% CI: 1.28–3.30) for those consuming >48 g daily. Among female past smokers, there was no association between alcohol consumption and pancreatitis risk at any dose (<24 g of alcohol daily (HR: 0.97; 95% CI: 0.77–1.23) and ⩾24 g daily (HR: 1.04; 95% CI: 0.61–1.76)). However, among current female smokers, consuming ⩾24 g of alcohol daily was associated with an increased risk of pancreatitis (HR: 1.67; 95% CI: 1.03–2.71), while those consuming <24 g had a marginally increased risk (HR: 1.32; 95% CI: 0.99–1.75).
Dose–response relationship
Greater cumulative exposure to smoke combined with alcohol use significantly increases the risk of AP compared to alcohol use alone. While low-intensity smoking combined with alcohol also elevates the risk, the impact is less pronounced compared to high-intensity smoking. For instance, in the study by Sadr-Azodi et al., 11 former smokers with ⩾20 pack-years who consumed alcohol showed a RR of 3.96 (95% CI: 1.87–8.39) compared to a non-significant association (RR: 1.97, 95% CI: 0.91–4.25) among former smokers with <20 pack-years who also consumed alcohol. Similar trends were observed in current smokers in the same study (⩾20 pack-years (RR: 4.12, 95% CI: 1.98–8.59) and <20 pack-years (RR: 2.13, 95% CI: 0.84–5.40)). 11 Likewise, among alcohol consumers reported by Morton et al., 27 former daily smokers who smoked ⩾1 pack/day (RR: 11.40, 95% CI: 6.47–20.10) showed strong association with AP than former smokers with <1 pack/day (RR: 5.80, 95% CI: 3.23–10.41). Compared to individuals who use neither substance, current smokers who are medium-risk alcohol consumers faced significantly increased risk (RR: 2.48, 95% CI: 1.52–4.05). The association was stronger for current smokers with high-risk drinking (RR: 4.85, 95% CI: 2.66–8.83; Figure 2). 26 No study examined the dose–response association of alcohol with AP risk among smokers.
Dose–response associations with smoking were also evident among studies on the risk of CP. For instance, Yadav et al. 30 found that smokers with >35 pack-years with heavy/very heavy alcohol intake showed an OR of 13.41 (95% CI: 5.23–34.39), compared to an OR of 4.47 (95% CI: 2.93–19.65) for smokers with moderate intensity (12–35 pack-years), and an OR of 3.13 (95% CI: 1.25–7.86) for light smokers (<12 pack-years) who drank at the same intensity. Similarly, among moderate drinkers, the OR increased with increasing smoking intensity: <12 pack years (OR: 1.87, 95% CI: 0.82–4.27), 12–35 pack-years (OR: 2.17, 95% CI: 0.96–4.90), and >35 pack-years (OR: 7.59, 95% CI: 2.93–19.65; Figure 3). 30
Quality assessment
Using the JBI appraisal tool, 2 studies were categorized as moderate quality,19,24 while the remaining 10 included studies were categorized as high quality. Yet, some concerns were identified with respect to validity.
Moderate-quality studies, like Ramzan’s case–control study, 24 relied on patient interviews for exposure data after pancreatitis development, increasing the risk of recall bias. In addition, exposures like smoking and alcohol lacked clarity regarding standardization, validity, and reliability, while the case definition of AP was insufficiently detailed. Similarly, Lowenfels’ retrospective case–control study 19 demonstrated inconsistent criteria for identifying CP cases (patients with alcoholic pancreatitis) and controls (patients with alcoholic liver cirrhosis), with unclear definitions and timing of exposure. Furthermore, in Lowenfels’ study, 19 cases and controls were not appropriately matched, likely introducing selection bias. In Ramzan’s study, 24 stratified analyses by age, gender, number of cigarettes/day, alcohol use, and residence were conducted, and univariate analysis used chi-square tests to obtain ORs and p-values to identify significant differences. However, residual confounding may remain without a multivariable analysis. Moderate-quality studies produced weaker and less stable estimates compared to high-quality studies. Among high-quality studies, some quality concerns were identified. For instance, some studies28,29,31 failed to adjust for both age and sex, which could result in biased estimates, overestimated or underestimated. Setiawan’s cohort study 31 did not clearly describe follow-up completeness or explore reasons for loss to follow-up. Similarly, the study by Haber et al. 18 did not clearly mention whether the exposure period of interest was long enough to be meaningful. Haber did not specify the study period. 18
Publication bias
Our visual inspection of funnel plots revealed a pattern suggestive of heterogeneity rather than publication bias for studies examining alcohol and smoking versus alcohol-only use (Supplemental Figure 1). This was expected given that studies had varying definitions and thresholds for substance use. The publication bias tests were not attempted as the sample size was underpowered to reliably detect small-study effects. In addition, no funnel plot was created for AP studies on smoking and alcohol versus neither substance use, as this consisted of only two studies with RR from 1.08 to 7.45.26,28
Discussion
This systematic review provides compelling evidence that combined alcohol and tobacco use increases the risk of either AP or CP compared to single substance use. Where examined, a dose-dependent relationship was observed among studies investigating multiple levels of smoking, with higher smoking contributing to the greater risk of pancreatitis. For AP, former and current smokers with high smoking intensity who also use alcohol showed a significantly increased risk than exclusive alcohol drinkers. Compared to people who use neither substance, the highest AP risk was observed among current smokers with high alcohol intake. Similarly, for CP, the greatest risks were found for the highest smoking category compared to non-smokers within each category of drinking. When comparing smoking and alcohol use versus neither, CP risk was the highest in those with both higher smoking and alcohol intensity. Interestingly, concurrent cannabis and alcohol use was associated with a significantly reduced risk of AP or CP compared to exclusive alcohol use, but only one study had examined this association. Although we observed high heterogeneity across studies, which likely stems from variation in exposure definitions, sample size, and study designs, our findings consistently highlight the additive effect of smoking and alcohol use on pancreatitis risk. Our exhaustive systematic review also identifies a gap in science on understanding the risk of pancreatitis with other combinations of substances, such as cannabis, amphetamines, and cocaine, behaviors that co-occur with alcohol and smoking, and also whether the association of polysubstance use with pancreatitis risk varies by sex or gender.
A prior systematic review on tobacco-only use versus never use 8 reported a significant increase in pancreatitis (AP, CP, and combined AP/CP) risk across all smoking statuses (current, former, and ever). Regarding the dose-dependent relationship for tobacco-only use and AP risk, the prior review 8 reported a RR of 1.13 (95% CI: 1.08–1.17) per 10 pack-years for current smokers compared to never smokers, with a stronger relationship observed between 0 and 20 pack-years, despite a nonlinear association overall. For former smokers, the prior review found a RR of 1.12 (95% CI: 1.07–1.17) per 10 pack-years, with a linear relationship. Furthermore, a significant linear relationship was found in CP, 8 with a RR of 1.22 (95% CI: 1.11–1.33) per 10 pack-years for current smokers. These findings align with the additive effects observed in our review, where smoking amplified pancreatitis risk in those who use alcohol. For instance, both Sadrz-Azodi et al. 11 and Yadav et al. 30 reported higher RR according to more extensive smoking history as compared to non-smoking in persons who consumed alcohol. Currently, the only existing alcohol-focused systematic review available 3 did not delineate AP and CP risk separately due to limited data but reported a combined risk for both forms of pancreatitis along with evidence of a dose–response association. Further studies on the additive and dose-dependent risk of pancreatitis due to alcohol in the presence of other substance use are critically needed.
Substance use induces pancreatitis through several pathophysiological mechanisms. Alcohol has long been the most extensively studied risk factor, but more recently, smoking has also been recognized as an equally prevalent risk factor for pancreatitis in epidemiological research. 32 Alcohol induces pancreatitis through calcium-mediated mitochondrial dysfunction and cell death, premature activation of trypsinogen within pancreatic acinar cells and macrophages, endoplasmic reticulum stress, impaired unfolded protein response, and impaired autophagy. 33 Smoking similarly impacts various pancreatic cell types, including acinar, ductal, and stellate cells.14,32 Smoking can induce pathological calcium signaling, oxidative stress, inhibit the cystic fibrosis transmembrane regulator, decrease fluid and bicarbonate secretion, and stimulate pancreatic fibrosis.14,32 Smoking may modulate immune responses by influencing anti-inflammatory and pro-inflammatory cytokine pathways. 34 In addition, smoking may impair oxidative energy metabolism and cause mitochondrial dysfunction through thiamine deficiency, which disrupts cellular energy processes and worsens pancreatic health. 35 Tobacco introduces toxins like nicotine and NNK (4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone), a tobacco-specific nitrosamine derived from nicotine. 34 NNK can trigger premature activation of digestive zymogens. 36
Furthermore, a study using mouse models, human primary acini, and the acinar cell line has demonstrated that alcohol and cigarette smoke exert synergistic effects. 14 Particularly, Lugea et al. 14 showed that smoking downregulates spliced X-box-binding protein 1 (XBP1), thereby impairing the adaptive unfolded protein response to alcohol. This disruption redirects the response to a maladaptive pathway characterized by upregulation of C/EBP homologous protein, ultimately resulting in unresolved endoplasmic reticulum stress and acinar cell death. Lugea et al. 14 also showed that combined alcohol and cigarette smoke exposure has worse effects on the pancreas (e.g., apoptosis and necrosis) compared to single exposure. Sahin-Toth et al. 32 build on Lugea et al.’s 14 findings by proposing the “multiple hits on multiple targets” model to describe how synergistic interactions among alcohol and cigarette smoke amplify harmful effects, highlighting the numerous signaling pathways that contribute to the development and progression of pancreatitis. Our findings provide epidemiological support for these studies,14,32 with consistent evidence for additive effects of alcohol and smoking.
Extending our findings to population-level data, a research letter based on an international cohort from 13 countries and 30 medical centers highlights the synergistic impact of smoking and alcohol on pancreatitis. 37 Individuals who engaged in both smoking and drinking developed their initial episode of AP 15 years prior to those who engaged in neither behavior. RAP was most common among people who smoked, regardless of alcohol use, whereas CP was most prevalent in those who engaged in both behaviors, suggesting a synergistic relationship and underscoring the role of smoking in disease progression. The study further demonstrated that smoking and alcohol each had a dose-dependent association with pancreatic tissue damage, and the combined use of both substances was linked to an increased prevalence of RAP and CP. 37 However, current mechanistic and epidemiological studies14,32,37 do not demonstrate multiplicative interaction of the two behaviors on pancreatitis risk.
Cannabis use has a complex relationship with pancreatitis. Cannabis has been linked to an increased risk of AP, with the risk rising as cannabis consumption duration extends. 38 In animal models, activation of the Cannabinoid (CB)-1 receptor has been shown to exacerbate disease activity. 39 However, CB1 and CB2 receptor agonists and selective CB2 receptor agonists have demonstrated positive effects, such as improving inflammation and tissue damage. 40 In CP, cannabinoid modulation is associated with the deactivation of pro-inflammatory pathways. 41 Our findings show a lower risk of AP and CP in those who use alcohol and cannabis compared to those reporting alcohol-only use may be due to the anti-inflammatory effects of cannabis via CB2 receptors, potentially offsetting the pro-inflammatory effects of alcohol. It is also possible that cannabis use dampens neuropathic pain, masking an underlying pancreatitis injury. 42
This systematic review has some notable strengths. To the best of our knowledge, it is the first to examine multiple substance use and the risk of pancreatitis, whereas previous reviews have primarily focused on single substance use such as tobacco or alcohol.3,8 Our study search considered multiple combinations of substance use including alcohol, tobacco, cannabis, and other illicit substances, providing a wide range of substances likely to be associated with pancreatitis risk. We also evaluated the risk of AP, CP, and combined pancreatitis, which gives a thorough understanding of the pancreatitis disease continuum. Furthermore, the use of an advanced AI tool for title and abstract review enhanced the efficiency of the review process. Finally, most of our included studies were categorized as having high quality.
Despite the above-mentioned strengths, we identified some limitations. Variability in exposure definitions across studies allowed for detailed insights into substance use patterns but resulted in increased heterogeneity. This limited our ability to estimate overall summary risk estimates or conduct standardized comparisons across studies. Consequently, we refrained from performing a meta-analysis to avoid generating unreliable pooled estimates. Furthermore, our findings are limited to non-gallstone-related pancreatitis to highlight the impact of substance use and minimize confounding effects from gallstone-induced cases. However, non-gallstone pancreatitis itself encompasses a range of other etiologies such as hypertriglyceridemia, hypercalcemia, and drug-induced causes, which may confound associations with alcohol, smoking, and marijuana. Genetic mutations like PRSS1, SPINK1, and CFTR also play a role. These alternative causes were not examined in the current study and represent an additional limitation. We do not completely rule out the possibility of publication bias, despite that our small study size within the meta-analysis diminished our ability to meaningfully assess this. We also identified some methodological limitations in the included studies, which have been outlined in the quality review section. Lack of adjustment for key confounders such as age and sex may have led to higher RR estimates, as both age and male sex are associated with more substance use and increased pancreatitis risk. Recall bias may have caused greater misclassification among controls in case–control studies, leading to higher RRs.
To date, research has predominantly emphasized single-substance use rather than multiple-substance use in relation to pancreatitis risk. Although we conducted a comprehensive search for multiple-substance use combinations, most included studies focused on smoking and alcohol co-use versus alcohol-only use, which leaves gaps in the understanding of other substance use combinations. Moreover, further research is needed in the following areas: (a) combined cannabis and alcohol use compared to alcohol-only use, where we found that only one included study showed reduced pancreatitis risk and (b) smoking and alcohol use compared to smoking alone, which showed non-significant risks for AP or CP. Notable gaps included no study on cannabis and smoking compared to smoking only, nor on concomitant recreational substance use along with alcohol.
Conclusion
This systematic review shows that combined alcohol and smoking significantly increases pancreatitis risk compared to single-substance use and also demonstrates a dose-dependent impact of smoking on pancreatitis among alcohol consumers. Evidence for the added impact of alcohol or cannabis among smokers remains insufficient. Pancreatitis patients require tailored substance use interventions due to the burden of pancreatitis-related hospitalizations and the negative effects of substance use. Cessation efforts in this population have often concentrated on quitting single-substance use, particularly alcohol. Therefore, more emphasis should be placed on addressing multiple-substance use and heavy use.
Supplemental Material
Acknowledgments
We appreciate Richard L. Wallace for assisting in creating search strategies.
Footnotes
Footnote Group
Contributor Information
Esther A. Adeniran, Department of Biomedical Sciences, Cedars-Sinai Medical Center, 6500 Wilshire Blvd, 15th Floor, 1544B, Los Angeles, CA 90048, USA.
Yi Jiang, Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA.
Dhiraj Yadav, Division of Gastroenterology, Hepatology and Nutrition, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA.
Judy Tan, Department of Biomedical Sciences, Cedars-Sinai Medical Center, Los Angeles, CA, USA.
Samuel Han, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, USA.
Simon K. Lo, Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Stephen J. Pandol, Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Christie Y. Jeon, Department of Biomedical Sciences, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Declarations
Untitled section
Footnote Group
References
Untitled section
References
- 1. Sankaran SJ, Xiao AY, Wu LM, et al. Frequency of progression from acute to chronic pancreatitis and risk factors: a meta-analysis. Gastroenterology 2015; 149: 1490–1500.e1.
- 2. Xiao AY, Tan ML, Wu LM, et al. Global incidence and mortality of pancreatic diseases: a systematic review, meta-analysis, and meta-regression of population-based cohort studies. Lancet Gastroenterol Hepatol 2016; 1: 45–55.
- 3. Irving HM, Samokhvalov AV, Rehm J. Alcohol as a risk factor for pancreatitis. A systematic review and meta-analysis. JOP 2009; 10: 387–392.
- 4. Quinlan JD. Acute pancreatitis. Am Fam Physician 2014; 90: 632–639.
- 5. Peery AF, Crockett SD, Murphy CC, et al. Burden and cost of gastrointestinal, liver, and pancreatic diseases in the United States: update 2021. Gastroenterology 2022; 162: 621–644.
- 6. Ouyang G, Pan G, Liu Q, et al. The global, regional, and national burden of pancreatitis in 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. BMC Med 2020; 18: 388.
- 7. Spanier BW, Dijkgraaf MG, Bruno MJ. Epidemiology, aetiology and outcome of acute and chronic pancreatitis: an update. Best Pract Res Clin Gastroenterol 2008; 22: 45–63.
- 8. Aune D, Mahamat-Saleh Y, Norat T, et al. Tobacco smoking and the risk of pancreatitis: a systematic review and meta-analysis of prospective studies. Pancreatology 2019; 19: 1009–1022.
- 9. Boileau-Falardeau M, Contreras G, Gariepy G, et al. Patterns and motivations of polysubstance use: a rapid review of the qualitative evidence. Health Promot Chronic Dis Prev Can 2022; 42: 47–59.
- 10. Adejumo AC, Akanbi O, Adejumo KL, et al. Reduced risk of alcohol-induced pancreatitis with cannabis use. Alcohol Clin Exp Res 2019; 43: 277–286.
- 11. Sadr-Azodi O, Andren-Sandberg A, Orsini N, et al. Cigarette smoking, smoking cessation and acute pancreatitis: a prospective population-based study. Gut 2012; 61: 262–267.
- 12. Atzendorf J, Rauschert C, Seitz NN, et al. The use of alcohol, tobacco, illegal drugs and medicines: an estimate of consumption and substance-related disorders in Germany. Dtsch Arztebl Int 2019; 116: 577–584.
- 13. Roche DJO, Bujarski S, Green R, et al. Alcohol, tobacco, and marijuana consumption is associated with increased odds of same-day substance co- and tri-use. Drug Alcohol Depend 2019; 200: 40–49.
- 14. Lugea A, Gerloff A, Su HY, et al. The combination of alcohol and cigarette smoke induces endoplasmic reticulum stress and cell death in pancreatic acinar cells. Gastroenterology 2017; 153: 1674–1686.
- 15. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021; 372: n71.
- 16. van Dijk SHB, Brusse-Keizer MGJ, Bucsan CC, et al. Artificial intelligence in systematic reviews: promising when appropriately used. BMJ Open 2023; 13: e072254.
- 17. Morgan RL, Whaley P, Thayer KA, et al. Identifying the PECO: a framework for formulating good questions to explore the association of environmental and other exposures with health outcomes. Environ Int 2018; 121: 1027–1031.
- 18. Haber PS, Wilson JS, Pirola RC. Smoking and alcoholic pancreatitis. Pancreas 1993; 8: 568–572.
- 19. Lowenfels AB, Zwemer FL, Jhangiani S, et al. Pancreatitis in a native American Indian population. Pancreas 1987; 2: 694–697.
- 20. Mersha AG, Gould GS, Bovill M, et al. Barriers and facilitators of adherence to nicotine replacement therapy: a systematic review and analysis using the capability, opportunity, motivation, and behaviour (COM-B) model. Int J Environ Res Public Health 2020; 17(23): 8895.
- 21. Tramel W, Schram B, Canetti E, et al. An examination of subjective and objective measures of stress in tactical populations: a scoping review. Healthcare (Basel) 2023; 11: 2515.
- 22. R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2022.
- 23. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21: 1539–1558.
- 24. Ramzan R, Ashraf A, Hassaan M, et al. Association of smoking and alcohol with acute pancreatitis in patients presenting in Tertiary Care Hospital Lahore. Pak J Med Health Sci 2022; 16: 227–229.
- 25. Ahmed Ali U, Issa Y, Hagenaars JC, et al. Risk of recurrent pancreatitis and progression to chronic pancreatitis after a first episode of acute pancreatitis. Clin Gastroenterol Hepatol 2016; 14: 738–746.
- 26. Lindkvist B, Appelros S, Manjer J, et al. A prospective cohort study of smoking in acute pancreatitis. Pancreatology 2008; 8: 63–70.
- 27. Morton C, Klatsky AL, Udaltsova N. Smoking, coffee, and pancreatitis. Am J Gastroenterol 2004; 99: 731–738.
- 28. Munigala S, Conwell DL, Gelrud A, et al. Heavy smoking is associated with lower age at first episode of acute pancreatitis and a higher risk of recurrence. Pancreas 2015; 44: 876–881.
- 29. Lin Y, Tamakoshi A, Hayakawa T, et al. Cigarette smoking as a risk factor for chronic pancreatitis: a case-control study in Japan. Research Committee on Intractable Pancreatic Diseases. Pancreas 2000; 21: 109–114.
- 30. Yadav D, Hawes RH, Brand RE, et al. Alcohol consumption, cigarette smoking, and the risk of recurrent acute and chronic pancreatitis. Arch Intern Med 2009; 169: 1035–1045.
- 31. Setiawan VW, Pandol SJ, Porcel J, et al. Prospective study of alcohol drinking, smoking, and pancreatitis: the multiethnic cohort. Pancreas 2016; 45: 819–825.
- 32. Sahin-Toth M, Hegyi P. Smoking and drinking synergize in pancreatitis: multiple hits on multiple targets. Gastroenterology 2017; 153: 1479–1481.
- 33. Lee PJ, Papachristou GI. New insights into acute pancreatitis. Nat Rev Gastroenterol Hepatol 2019; 16: 479–496.
- 34. Thrower E. Pathologic cellular events in smoking-related pancreatitis. Cancers (Basel) 2015; 7: 723–735.
- 35. Srinivasan P, Subramanian VS, Said HM. Effect of the cigarette smoke component, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), on physiological and molecular parameters of thiamin uptake by pancreatic acinar cells. PLoS One 2013; 8: e78853.
- 36. Alexandre M, Uduman AK, Minervini S, et al. Tobacco carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone initiates and enhances pancreatitis responses. Am J Physiol Gastrointest Liver Physiol 2012; 303: G696–G704.
- 37. Szentesi A, Farkas N, Sipos Z, et al. Alcohol consumption and smoking dose-dependently and synergistically worsen local pancreas damage. Gut 2022; 71: 2601–2602.
- 38. Han B, Gfroerer JC, Colliver JD. Associations between duration of illicit drug use and health conditions: results from the 2005–2007 national surveys on drug use and health. Ann Epidemiol 2010; 20: 289–297.
- 39. Dembinski A, Warzecha Z, Ceranowicz P, et al. Cannabinoids in acute gastric damage and pancreatitis. J Physiol Pharmacol 2006; 57(Suppl. 5): 137–154.
- 40. Michalski CW, Laukert T, Sauliunaite D, et al. Cannabinoids ameliorate pain and reduce disease pathology in cerulein-induced acute pancreatitis. Gastroenterology 2007; 132: 1968–1978.
- 41. Utomo WK, de Vries M, Braat H, et al. Modulation of human peripheral blood mononuclear cell signaling by medicinal cannabinoids. Front Mol Neurosci 2017; 10: 14.
- 42. Mucke M, Phillips T, Radbruch L, et al. Cannabis-based medicines for chronic neuropathic pain in adults. Cochrane Database Syst Rev 2018; 3: CD012182.