Key PointsQuestion
Is nonfasting mild-to-moderate hypertriglyceridemia associated with acute pancreatitis?
Findings
In 116 550 individuals from the general population, nonfasting mild-to-moderate hypertriglyceridemia of 177 mg/dL (2 mmol/L) or higher was associated with high risk of acute pancreatitis, with risk estimates higher than for myocardial infarction.
Meaning
Mild-to-moderate hypertriglyceridemia is associated with increased risk not only for myocardial infarction but also for acute pancreatitis.
Importance
Severe hypertriglyceridemia is associated with increased risk of acute pancreatitis. However, the threshold above which triglycerides are associated with acute pancreatitis is unclear.
Objective
To test the hypothesis that nonfasting mild-to-moderate hypertriglyceridemia (177-885 mg/dL; 2-10 mmol/L) is also associated with acute pancreatitis.
Design, Setting, and Participants
This prospective cohort study examines individuals from the Copenhagen General Population Study in 2003 to 2015 and the Copenhagen City Heart Study initiated in 1976 to 1978 with follow-up examinations in 1981 to1983, 1991 to 1994, and in 2001 to 2003. Median follow-up was 6.7 years (interquartile range, 4.0-9.4 years); and includes 116 550 individuals with a triglyceride measurement from the Copenhagen General Population Study (n = 98 649) and the Copenhagen City Heart Study (n = 17 901). All individuals were followed until the occurrence of an event, death, emigration, or end of follow-up (November 2014), whichever came first.
Exposures
Plasma levels of nonfasting triglycerides.
Main Outcomes and Measures
Hazard ratios (HRs) for acute pancreatitis (n = 434) and myocardial infarction (n = 3942).
Results
Overall, 116 550 individuals were included in this study (median [interquartile range] age, 57 [47-66] years). Compared with individuals with plasma triglyceride levels less than 89 mg/dL (<1 mmol/L), the multivariable adjusted HRs for acute pancreatitis were 1.6 (95% CI, 1.0-2.6; 4.3 events/10 000 person-years) for individuals with triglyceride levels of 89 mg/dL to 176 mg/dL (1.00 mmol/L-1.99 mmol/L), 2.3 (95% CI, 1.3-4.0; 5.5 events/10 000 person-years) for 177 mg/dL to 265 mg/dL (2.00 mmol/L-2.99 mmol/L), 2.9 (95% CI, 1.4-5.9; 6.3 events/10 000 person-years) for 366 mg/dL to 353 mg/dL (3.00 mmol/L-3.99 mmol/L), 3.9 (95% CI, 1.5-10.0; 7.5 events/10 000 person-years) for 354 mg/dL-442 mg/dL (4.00 mmol/L-4.99 mmol/L), and 8.7 (95% CI, 3.7-20.0; 12 events/10 000 person-years) for individuals with triglyceride levels greater than or equal to 443 mg/dL (≥5.00 mmol/L) (trend, P = 6 × 10−8). Corresponding HRs for myocardial infarction were 1.6 (95% CI, 1.4-1.9; 41 events/10 000 person-years), 2.2 (95% CI, 1.9-2.7; 57 events/10 000 person-years), 3.2 (95% CI, 2.6-4.1; 72 events/10 000 person-years), 2.8 (95% CI, 2.0-3.9; 68 events/10 000 person-years), and 3.4 (95% CI, 2.4-4.7; 78 events/10 000 person-years) (trend, P = 6 × 10−31), respectively. The multivariable adjusted HR for acute pancreatitis was 1.17 (95% CI, 1.10-1.24) per 89 mg/dL (1 mmol/L) higher triglycerides. When stratified by sex, age, education, smoking, hypertension, statin use, study cohort, diabetes, body mass index (calculated as weight in kilograms divided by height in meters squared), alcohol intake, and gallstone disease, these results were similar with no statistical evidence of interaction.
Conclusions and Relevance
Nonfasting mild-to-moderate hypertriglyceridemia from 177 mg/dL (2 mmol/L) and above is associated with high risk of acute pancreatitis, with HR estimates higher than for myocardial infarction.
Acute pancreatitis is a common and potentially fatal condition.1,2 The prevalence of acute pancreatitis is on the rise; in the United States, hospital admissions for acute pancreatitis increased from 4 events per 10 000 person-years in 1988 to 7 events per 10 000 person-years in 2002.3 Gallstone disease and high alcohol consumption are the leading causes of acute pancreatitis, while infections, drugs, endoscopic retrograde cholangiopancreatography, and metabolic conditions are more uncommon etiologies.4
Severe hypertriglyceridemia (>885 mg/dL; >10 mmol/L) is also a well-established cause of acute pancreatitis.5-10 However, there is no consensus on a clear threshold above which triglycerides are associated with acute pancreatitis.10,11 According to guidelines from the American College of Gastroenterology and the Endocrine Society, triglyceride levels greater than 1000 mg/dL (>11.3 mmol/L) should be considered a risk factor of acute pancreatitis12,13; the European Society of Cardiology and the European Atherosclerosis Society set this cut point at 885 mg/dL (10 mmol/L),14 and the National Cholesterol Education Program Adult Treatment Panel III recommends triglyceride-lowering therapy when triglycerides are greater than or equal to 500 mg/dL (≥5.65 mmol/L) to prevent acute pancreatitis.15
We tested the hypothesis that nonfasting mild-to-moderate hypertriglyceridemia between 177 mg/dL and 885 mg/dL (2-10 mmol/L) is also associated with high risk of acute pancreatitis. In addition, we examined the association between mild-to-moderate hypertriglyceridemia and risk of myocardial infarction16 for comparison of effect sizes, as well as between mild-to-moderate hypertriglyceridemia and levels of plasma lipase and pancreatic amylase for biochemical support of our findings. As lipase and pancreatic amylase are the standard blood test to diagnose acute pancreatitis, high levels of these biomarkers at mild-to-moderate hypertriglyceridemia may indicate subclinical acute pancreatitis.
We studied white individuals of Danish descent from the Danish general population included in 2 similar prospective studies, the Copenhagen City Heart Study and the Copenhagen General Population Study, selected randomly based on their Central Person Register number. The studies were approved by Herlev and Gentofte Hospital and a Danish ethical committee, and both studies complied with the Declaration of Helsinki.
The Copenhagen City Heart Study was initiated in 1976 through 1978 with follow-up examinations in 1981 to 1983, 1991 to 1994, and in 2001 to 2003.17 Data were obtained from a physical examination, blood samples, and a self-administered questionnaire reviewed by an examiner at the day of attendance. We included 17 901 consecutive individuals with a triglyceride measurement. The Copenhagen General Population Study recruited in 2003 through 2015, with an examination similar to the Copenhagen City Heart Study. We included 98 649 consecutive individuals with a triglyceride measurement. The combined study included 116 550 individuals, in whom, during a median follow-up of 6.7 years (interquartile range, 4.0-9.4 years), 434 individuals developed acute pancreatitis and 3942 had a myocardial infarction.
All individuals were followed until the occurrence of an event, death (n = 17 331), emigration (n = 598), or end of follow-up (November 14, 2014), whichever came first; we did not lose track of any individuals. Information regarding diagnosis of acute pancreatitis (World Health Organization International Classification of Diseases 8th Edition (ICD-8), codes 577.00, 577.01, 577.08, and 577.09; ICD-10 code K85), chronic pancreatitis (ICD-8 codes 577.10, 577.11, and 577.19; ICD-10 code K86) and myocardial infarction (ICD-8 code 410; ICD-10 codes I21 and I22) was collected by reviewing all hospital diagnoses and all deaths entered in the national Danish Patient Registry and the national Danish Causes of Death Registry. We included myocardial infarction as a positive control16 to assess the validity of the tests performed and for comparison of effect sizes.
Laboratory Analysis and Covariates
Plasma levels of nonfasting triglycerides (n = 116 550), lipase (n = 15 856), and pancreatic amylase (n = 92 672) were measured on fresh blood samples with use of standard hospital assays. Blood sampling was not related to an event of acute pancreatitis. Nonfasting triglycerides are maximally increased by 26 mg/dL (0.3 mmol/L) after intake of habitual meals compared with fasting triglycerides.18-24 Lipase and pancreatic amylase were only measured in the Copenhagen General Population Study: 10 085 individuals had measurement of both lipase and pancreatic amylase, while 206 individuals had no measurement of either lipase or pancreatic amylase. For lipase, no interference is observed by lipemia up to 1000 mg/dL (11.3 mmol/L) of triglycerides and for pancreatic amylase up to 2000 mg/dL (22.6 mmol/L) of triglycerides (lipase and pancreatic amylase assays; DiaSys).
All patient information on age and sex was complete. Information on other covariates was 99.4% complete. The missing values were imputed based on age and sex; however, in analysis only including individuals with complete data, results were similar to those reported. Assessment of covariates is described in eMethods in the Supplement.
We used Stata/SE13.1(StataCorp). Individuals were classified into 6 clinical categories of triglycerides (<89, 89-176, 177-265, 266-353, 354-442, and ≥443 mg/dL; <1.00, 1.00-1.99, 2.00-2.99, 3.00-3.99, 4.00-4.99, and ≥5.00 mmol/L), as done previously.17 For trend tests using Cuzick nonparametric test or within Cox regression models, these categories were coded 0, 1, 2, 3, 4, and 5 for higher levels of triglycerides.
Hazard ratios (HRs) for acute pancreatitis and myocardial infarction were estimated using Cox proportional hazard regressions models. Age was the time scale using left truncation (delayed entry), which implies automatic adjustment. We excluded individuals with events before the examination day: for acute pancreatitis we excluded 506 individuals, leaving 116 044 for analyses, and for myocardial infarction we excluded 2402, leaving 114 148 for analyses. Baseline for all included individuals was the date triglycerides were measured.
Cox regression models were adjusted for age and sex, or for age, sex, and baseline markers of lifestyle and known risk factors for pancreatitis (ie, for education, smoking, hypertension, statin use, study cohort, and birth year). We fitted 4 additional multivariable adjusted models: 1 further adjusted for alcohol intake and one for diabetes and body mass index (BMI; calculated as weight in kilograms divided by height in meters squared), as these may be on the biological pathway between hypertriglyceridemia and acute pancreatitis, 1 further adjusted for time since last meal, and 1 further adjusted for high-density lipoprotein cholesterol. We calculated stratum-specific HRs for acute pancreatitis by performing subgroup analyses according to diabetes, BMI (<25 vs ≥25), alcohol intake (<8 vs ≥8 units/week), and gallstone disease.
Because of random measurement error in the measurement of plasma triglycerides values at baseline, the true association between triglycerides and acute pancreatitis or myocardial infarction is attenuated; this statistical phenomenon is known as regression dilution bias.25,26 Therefore, HRs and CIs were corrected for regression dilution bias using a nonparametric method,27 as done previously.17,28 For this correction, we used triglycerides values from 4723 individuals from the Copenhagen City Heart Study who attended both the 1991 to 1994 examination and the 2001 to 2003 examination. A regression dilution ratio of 0.57 was calculated for triglycerides.
Hazard ratios for acute pancreatitis were also calculated per 89 mg/dL (1 mmol/L) higher levels of triglycerides. To assess the risk of acute pancreatitis in different subgroups, we again calculated stratum-specific HRs for acute pancreatitis by performing subgroup analyses according to age, sex, education, smoking, hypertension, statin use, study cohort, diabetes, BMI (<25 vs ≥25 kg/m2), alcohol intake (women, <14 [low]/≥14 [high] and men, <21 [low]/≥21 [high]), and gallstone disease. Tests for interaction were performed by introducing a 2-factor interaction term in the Cox regression model. For the test of interaction of age with triglyceride levels on acute pancreatitis risk, we used years of follow-up as the time scale. Also, we performed subgroup analyses by only allowing individuals to enter the Cox regression model at 1, 2, 3, 4, and 5 years after measurement of triglycerides.
Lipase and pancreatic amylase, were log2-transformed to approximate normal distributions. The association between categories of triglycerides and levels of lipase and pancreatic amylase were examined by general linear models, adjusted for age and sex or multivariable adjusted for age, sex, education, smoking, hypertension, statin use, and birth year. Afterwards geometric mean for lipase and pancreatic amylase were calculated.
Baseline characteristics of 116 550 individuals from the general population by clinical categories of triglyceride levels are shown in the Table.
Acute Pancreatitis and Myocardial Infarction
The risk of acute pancreatitis and myocardial infarction was higher as a function of higher levels of plasma triglycerides (Figure 1). Compared with individuals with plasma triglyceride levels less than 89 mg/dL (<1 mmol/L), the multivariable adjusted HRs for acute pancreatitis were 1.6 (95% CI, 1.0-2.6; 4.3 events/10 000 person-years) for individuals with triglyceride levels of 89 mg/dL to 176 mg/dL (1.00 mmol/L-1.99 mmol/L), 2.3 (95% CI, 1.3-4.0; 5.5 events/10 000 person-years) for 177 mg/dL to 265 mg/dL (2.00 mmol/L-2.99 mmol/L), 2.9 (95% CI, 1.4-5.9; 6.3 events/10 000 person-years) for 366 mg/dL to 353 mg/dL (3.00 mmol/L-3.99 mmol/L), 3.9 (95% CI, 1.5-10; 7.5 events/10 000 person-years) for 354 mg/dL to 442 mg/dL (4.00 mmol/L-4.99 mmol/L), and 8.7 (95% CI, 3.7-20.0; 12 events/10 000 person-years) for individuals with triglyceride levels greater than or equal to 443 mg/dL (≥5.00 mmol/L) (trend, P = 6 × 10−8). Corresponding HRs for myocardial infarction were 1.6 (95% CI, 1.4-1.9; 41 events/10 000 person-years), 2.2 (95% CI, 1.9-2.7; 57 events/10 000 person-years), 3.2 (95% CI, 2.6-4.1; 72 events/10 000 person-years), 2.8 (95% CI, 2.0-3.9; 68 events/10 000 person-years), and 3.4 (95% CI, 2.4-4.7; 78 events/10 000 person-years) (trend, P = 6 × 10−31), respectively. See eFigure 1 in the Supplement for results before correction for regression dilution bias.
After additional adjustment for alcohol intake, the risk estimates were similar, while additional adjustment for diabetes and BMI did attenuate risk estimates. Results were similar when further adjusted for time since last meal (eFigure 2 in the Supplement), after additional exclusion of individuals with chronic pancreatitis before baseline (n = 140) (eFigure 3 in the Supplement), and in separate analyses of data from the Copenhagen City Heart Study and the Copenhagen General Population Study (eFigure 4 in the Supplement). Results for acute pancreatitis were similar when further adjusted for high-density lipoprotein cholesterol, while the association between triglycerides and myocardial infarction were attenuated (eFigure 5 in the Supplement). We found similar results in individuals with and without diabetes (interaction, P = .83), with BMI less than 25 and greater than or equal to 25 (P = .92), alcohol intake less than 8 units per week vs 8 or more units per week (P = .93), and in those with or without gallstone disease (P = .40) (eFigure 6 in the Supplement).
When plasma triglycerides were expressed on a continuously scale, the multivariable adjusted HR for acute pancreatitis was 1.17 (95% CI, 1.10-1.24) per 89 mg/dL (1 mmol/L) higher triglycerides (Figure 2). When stratified by sex, age, education, smoking, hypertension, statin use, study cohort, diabetes, BMI, alcohol intake, and gallstone disease, the association between higher levels of plasma triglycerides and risk of acute pancreatitis was still significant in most strata (Figure 2). We found no statistical evidence of interaction in any of the analyzed strata. See eFigure 7 in the Supplement for results before correction for regression dilution bias. To exclude reverse causation, that is, that acute pancreatitis could cause elevated triglycerides, we also performed subgroup analyses by only allowing individuals to enter the Cox regression model at 1, 2, 3, 4, and 5 years after the initial start of follow-up and date of triglyceride measurement: the HRs for acute pancreatitis per 89 mg/dL (1 mmol/L) increase in triglycerides were similar throughout (eFigure 8 in the Supplement).
Absolute Risk for Acute Pancreatitis and Myocardial Infarction
For acute pancreatitis, the absolute risk from triglyceride levels less than 89 mg/dL (<1 mmol/L) to 89 mg/dL to 176 mg/dL (1.00-1.99 mmol/L) was higher by 1.6 events/10 000 person-years; to 177 mg/dL to 265 mg/dL (2.00 mmol/L-2.99 mmol/L) by 2.8 events/10 000 person-years; to 266 mg/dL to 353 mg/dL (3.00 mmol/L-3.99 mmol/L) by 3.6 events/10 000 person-years; to 354 mg/dL to 442 mg/dL (4.00 mmol/L-4.99 mmol/L) by 4.8 events/10 000 person-years, and to triglyceride levels greater than or equal to 443 mg/dL (≥5 mmol/L) it was higher by 9.3 events/10 000 person-years (Figure 1). For myocardial infarction the corresponding higher absolute risks were 19, 35, 50, 46, and 56 events/10 000 person-years, respectively.
Association Between Triglycerides and Lipase and Pancreatic Amylase
Baseline characteristics of individuals are shown as a function of plasma lipase or pancreatic amylase tertiles in the eTable in the Supplement.
Higher levels of plasma triglycerides were associated with higher levels of plasma lipase and with lower levels of pancreatic amylase (Figure 3). The multivariable adjusted geometric mean values for lipase were 28.1 units/L for triglyceride levels of less than 89 mg/dL (<1 mmol/L), 29.2 units/L for 89 mg/dL to 176 mg/dL (1.00 mmol/L-1.99 mmol/L), 31.0 units/L for 177 mg/dL to 265 mg/dL (2.00-2.99 mmol/L), 32.6 units/L for 266 mg/dL to 353 mg/dL (3.00 mmol/L-3.99 mmol/L), 32.7 units/L for 354 mg/dL to 442 mg/dL (4.00 mmol/L-4.99 mmol/L), and 34.0 units/L for triglyceride levels greater than or equal to 443 mg/dL (≥5.00 mmol/L) (trend, P = 2 × 10−23) (Figure 3C). Corresponding values for pancreatic amylase were 32.8 units/L, 32.4 units/L, 31.8 units/L, 31.2 units/L, 31.2 units/L, and 29.8 units/L, respectively (trend, P = 2 × 10−57) (Figure 3D). Results were similar when we excluded individuals with triglycerides levels above the recommended levels for the assay (eFigure 9 in the Supplement).
We found that mild-to-moderate hypertriglyceridemia of 177 mg/dL (2 mmol/L) and above was associated with high risk of acute pancreatitis in the general population. Hazard ratios for acute pancreatitis were even higher than for myocardial infarction, the latter in accordance with previous findings.16 Notably, there was roughly 10 times more myocardial infarction than acute pancreatitis events, also illustrated by the higher absolute risk from triglyceride levels less than 89 mg/dL (<1 mmol/L) to levels greater than or equal to 443 mg/dL (≥5 mmol/L) of 9.3 events/10 000 person-years for acute pancreatitis and of 56 events/10 000 person-years for myocardial infarction The present data may have implications in changing the standard of care in patients with hypertriglyceridemia both with respect to acute pancreatitis and myocardial infarction.
The mechanism by which triglycerides lead to acute pancreatitis is not known. One theory is that at high levels, triglycerides are hydrolyzed by pancreatic lipase producing high levels of free fatty acids locally in the pancreas.29,30 These will self-aggregate to micellar structures with detergent properties, which in turn will damage acinar cells and the vascular bed inside the pancreas, causing ischemia and inflammation. Also, it has been proposed that plasma hyperviscosity due to high levels of chylomicrons (with high content of triglycerides) may lead to impaired blood flow, ischemia, and inflammation in the pancreas.11,31 In accordance, higher levels of triglycerides in rats with acute pancreatitis have been found to induce inflammatory mediators in the pancreas.32 Further in support, a Mendelian randomization study33 found genetically mild-to-moderately high remnant cholesterol (highly correlated with triglycerides) to be casually associated with high C-reactive protein levels, and because remnant cholesterol is highly correlated with triglycerides, this finding suggest that mild-to-moderate hypertriglyceridemia is associated with whole-body low-grade inflammation.
Lipoprotein lipase is the key enzyme degrading triglycerides in plasma,34 and mink with hyperlipidemia due to lipoprotein lipase deficiency (which leads to high levels of chylomicrons and triglycerides) nearly always develop pancreatitis.35 In humans, lipoprotein lipase deficiency leads to the familial hyperchylomicronemia syndrome with severe hypertriglyceridemia which often, as in mink, leads to acute pancreatitis.9,10,34 Interestingly, individuals with the familial chylomicronemia syndrome and severe hypertriglyceridemia seem to be protected against atherosclerotic cardiovascular disease,34,36 where the explanation might be that the very large triglyceride-rich lipoproteins cannot enter into the arterial intima.37 On the other hand, when mild-to-moderate hypertriglyceridemia is present, the medium-sized, triglyceride-rich lipoproteins can enter into the arterial intima where lipoprotein lipase will degrade triglycerides leading to liberation of free fatty acids.36 Free fatty acids might then induce intimal inflammation and thus might likely induce inflammation in atherosclerotic plaques that are possibly involved in further development of myocardial infarction, as confirmed in the present study to be associated with mild-to-moderate hypertriglyceridemia. It might be a similar mechanism that takes place in the pancreas, where the free fatty acids may lead to inflammation and thus pancreatitis.36 Interestingly, a meta-analysis of randomized controlled trials found that use of statins is associated with reduced risk of acute pancreatitis in patients with normal or mildly increased levels of triglycerides.38
When we adjusted for diabetes and BMI the association between mild-to-moderate hypertriglyceridemia and acute pancreatitis was attenuated, probably because both diabetes and obesity lead to higher levels of triglycerides,9,39 and therefore, are part of the biological pathway to acute pancreatitis. There are studies suggesting that diabetes itself, when controlling for hypertriglyceridemia, gallstones, etc, is associated with high risk of acute pancreatitis.40,41 Therefore, some might argue that our results might be explained by a high prevalence of diabetes and/or higher BMI rather than high triglycerides. However, because we found similar results in individuals with and without diabetes and with a BMI less than 25 and greater than or equal to 25, it seems unlikely that high prevalence of diabetes and high BMI can explain the association observed in the present study between mild-to-moderate hypertriglyceridemia and acute pancreatitis. Likewise, our stratified analyses also found it unlikely that gallstone disease should be the explanation behind our findings. Finally, because we obtained similar results even when we excluded individuals developing acute pancreatitis from 1 to 5 years after triglyceride measurement, it seems unlikely that reverse causality, that is, acute pancreatitis leading to increased levels of triglyceride, can explain the association observed between mild-to-moderate hypertriglyceridemia and acute pancreatitis.
In support of our findings, a Swedish prospective cohort study42 with 33 346 individuals found a HR for acute pancreatitis of 1.55 (95% CI, 1.09-2.21) for individuals in the fourth quartile (≥145 mg/dL; ≥1.64 mmol/L) vs the first quartile (≤75 mg/dL; ≤0.85 mmol/L) of triglycerides. Also, a population-based study from Scotland43 with 67 269 individuals, with triglyceride levels 150 mg/dL or less (≤1.7 mmol/L) as the reference, found HRs for acute pancreatitis of 1.50 (95% CI, 1.14-1.97) for individuals with triglyceride levels between 150 mg/dL and 499 mg/dL (1.7-5.6 mmol/L) and an HR of 3.2 (1.99-5.16) for triglycerides greater than or equal to 500 mg/dL (≥5.6 mmol/L). Using another approach, a retrospective cohort study44 of 41 210 individuals with hypertriglyceridemia at baseline found that lower levels of triglycerides during follow-up were associated with reductions in the incidence of acute pancreatitis. Notably, patients whose follow-up triglyceride levels were reduced to less than 200 mg/dL (<2.3 mmol/L) had the largest reduction in the incidence of acute pancreatitis.
In apparent contrast to our findings, and in support of the commonly held notion that only severe hypertriglyceridemia increases the risk of acute pancreatitis, Sandhu et al8 found in a cohort of 91 patients with triglyceride levels between 885 md/dL and 1770 mg/dL (10-20 mmol/L) that only 3 patients had a history of acute pancreatitis, and these 3 all had triglyceride levels greater than 1770 mg/dL (>20 mmol/L) at the time of acute pancreatitis.
Also, Linares et al7 showed that 85% of the patients with acute pancreatitis had maximal triglyceride levels greater than 3000 mg/dL (>34 mmol/L). Both of these retrospective studies only included individuals referred to lipid clinics and consequently only high-risk individuals. In contrast, we conducted a prospective cohort study in the general population. However, our findings may indicate that when a person has triglycerides of 177 mg/dL (2 mmol/L) or above, then that person is at increased risk of later experiencing episodes of even higher triglycerides, and only at such instances will acute pancreatitis develop.
Our finding that high levels of plasma triglycerides are associated with high levels of plasma lipase is consistent with that of Steinberg et al45 in patients with type 2 diabetes. Steinberg et al also found, in accordance with our results, that only lipase and not amylase was higher with higher levels of plasma triglycerides. The explanation for this difference might be that, compared with amylase, lipase remains elevated for a longer time after pancreatic injury, and lipase is not as affected by chronic pancreatic insufficiency as amylase46,47; indeed, chronic pancreatitis might explain the observed low levels of pancreatic amylase in individuals with mild-to-moderate hypertriglyceridemia. Alternatively, one might think that it reflects a combination of: (1) half-life of the enzymes; (2) time since last meal; and (3) interference of triglycerides with the lipase assay, because the lipase assay contains diglycerides and the pancreatic amylase assay does not.
Among the major strengths of our study are the large cohorts of the general population, the prospective population-based design, the large number of events, no losses to follow-up, and the possibility to correct for regression dilution bias because this correction helps avoid underestimation of risk estimates.25
A study limitation includes diagnostic misclassification. However, although the validity of acute pancreatitis diagnosis in the national Danish Patient Registry is estimated to be 82% (95% CI, 73%-89%),48 such misclassification is likely to be nondifferential to triglyceride levels and is therefore unlikely to explain the presents results. In addition, we were not able to retrieve workable information on subtypes of acute pancreatitis (alcohol-related, gallstone-related, etc) from the Danish registries, and for that reason we were not able to compare the association with hypertriglyceridemia in different subtypes of acute pancreatitis. Also, lipase and pancreatic amylase were only available in a subfraction of respectively 16% and 94% of all individuals; however, characteristics in those with and without these measurements were similar.
Mild-to-moderate hypertriglyceridemia at 177 mg/dL (2 mmol/L) and above is associated with high risk of acute pancreatitis in the general population, with HRs higher than for myocardial infarction. These observations might be useful for forthcoming revisions of guidelines on the management of hypertriglyceridemia.
Corresponding Author: Børge G. Nordestgaard, MD, DMSc, Professor, Chief Physician, Department of Clinical Biochemistry, Herlev and Gentofte Hospital, Copenhagen University Hospital, Herlev Ringvej 75, DK-2730 Herlev, Denmark (boerge.nordestgaard@regionh.dk).
Published Online: November 7, 2016. doi:10.1001/jamainternmed.2016.6875
Author Contributions: Dr Nordestgaard had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: All Authors.
Acquisition, analysis, or interpretation of data: All Authors.
Drafting of the manuscript: Pedersen.
Critical revision of the manuscript for important intellectual content: Langsted, Nordestgaard.
Statistical analysis: Pedersen, Langsted.
Obtained funding: Nordestgaard.
Conflict of Interest Disclosures: Dr Nordestgaard reports consultancies or talks sponsored by AstraZeneca, Merck, Omthera, Ionis, and Kowa. No other conflicts are reported.
Funding/Support: This study was supported by the Herlev and Gentofte Hospital and Copenhagen University Hospital.
Role of the Funder/Sponsor: The funders/sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
2.Peery
AF, Dellon
ES, Lund
J,
et al. Burden of gastrointestinal disease in the United States: 2012 update.
Gastroenterology. 2012;143(5):1179-87.e1, 3.
PubMedGoogle ScholarCrossref 3.Fagenholz
PJ, Castillo
CF, Harris
NS, Pelletier
AJ, Camargo
CA
Jr. Increasing United States hospital admissions for acute pancreatitis, 1988-2003.
Ann Epidemiol. 2007;17(7):491-497.
PubMedGoogle ScholarCrossref 5.Toskes
PP. Hyperlipidemic pancreatitis.
Gastroenterol Clin North Am. 1990;19(4):783-791.
PubMedGoogle Scholar 6.Fortson
MR, Freedman
SN, Webster
PD
III. Clinical assessment of hyperlipidemic pancreatitis.
Am J Gastroenterol. 1995;90(12):2134-2139.
PubMedGoogle Scholar 7.Lloret Linares
C, Pelletier
AL, Czernichow
S,
et al. Acute pancreatitis in a cohort of 129 patients referred for severe hypertriglyceridemia.
Pancreas. 2008;37(1):13-2.
PubMedGoogle ScholarCrossref 8.Sandhu
S, Al-Sarraf
A, Taraboanta
C, Frohlich
J, Francis
GA. Incidence of pancreatitis, secondary causes, and treatment of patients referred to a specialty lipid clinic with severe hypertriglyceridemia: a retrospective cohort study.
Lipids Health Dis. 2011;10:157.
PubMedGoogle ScholarCrossref 9.Hegele
RA, Ginsberg
HN, Chapman
MJ,
et al; European Atherosclerosis Society Consensus Panel. The polygenic nature of hypertriglyceridaemia: implications for definition, diagnosis, and management.
Lancet Diabetes Endocrinol. 2014;2(8):655-666.
PubMedGoogle ScholarCrossref 10.Scherer
J, Singh
VP, Pitchumoni
CS, Yadav
D. Issues in hypertriglyceridemic pancreatitis: an update.
J Clin Gastroenterol. 2014;48(3):195-203.
PubMedGoogle ScholarCrossref 11.Valdivielso
P, Ramírez-Bueno
A, Ewald
N. Current knowledge of hypertriglyceridemic pancreatitis.
Eur J Intern Med. 2014;25(8):689-694.
PubMedGoogle ScholarCrossref 12.Tenner
S, Baillie
J, DeWitt
J, Vege
SS; American College of Gastroenterology. American College of Gastroenterology guideline: management of acute pancreatitis.
Am J Gastroenterol. 2013;108(9):1400-1415, 1416.
PubMedGoogle ScholarCrossref 13.Berglund
L, Brunzell
JD, Goldberg
AC,
et al; Endocrine society. Evaluation and treatment of hypertriglyceridemia: an Endocrine Society clinical practice guideline.
J Clin Endocrinol Metab. 2012;97(9):2969-2989.
PubMedGoogle ScholarCrossref 14.Catapano
AL, Reiner
Z, De Backer
G,
et al; European Society of Cardiology (ESC); European Atherosclerosis Society (EAS). ESC/EAS Guidelines for the management of dyslipidaemias the Task Force for the management of dyslipidaemias of the European Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS).
Atherosclerosis. 2011;217(1):3-46.
PubMedGoogle ScholarCrossref 15.National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report.
Circulation. 2002;106(25):3143-3421.
PubMedGoogle Scholar 17.Nordestgaard
BG, Benn
M, Schnohr
P, Tybjaerg-Hansen
A. Nonfasting triglycerides and risk of myocardial infarction, ischemic heart disease, and death in men and women.
JAMA. 2007;298(3):299-308.
PubMedGoogle ScholarCrossref 18.Mora
S, Rifai
N, Buring
JE, Ridker
PM. Fasting compared with nonfasting lipids and apolipoproteins for predicting incident cardiovascular events.
Circulation. 2008;118(10):993-1001.
PubMedGoogle ScholarCrossref 19.Langsted
A, Freiberg
JJ, Nordestgaard
BG. Fasting and nonfasting lipid levels: influence of normal food intake on lipids, lipoproteins, apolipoproteins, and cardiovascular risk prediction.
Circulation. 2008;118(20):2047-2056.
PubMedGoogle ScholarCrossref 20.Steiner
MJ, Skinner
AC, Perrin
EM. Fasting might not be necessary before lipid screening: a nationally representative cross-sectional study.
Pediatrics. 2011;128(3):463-470.
PubMedGoogle Scholar 21.Langsted
A, Nordestgaard
BG. Nonfasting lipids, lipoproteins, and apolipoproteins in individuals with and without diabetes: 58 434 individuals from the Copenhagen General Population Study.
Clin Chem. 2011;57(3):482-489.
PubMedGoogle ScholarCrossref 22.Sidhu
D, Naugler
C. Fasting time and lipid levels in a community-based population: a cross-sectional study.
Arch Intern Med. 2012;172(22):1707-1710.
PubMedGoogle ScholarCrossref 24.Nordestgaard
BG, Langsted
A, Mora
S,
et al; European Atherosclerosis Society (EAS) and the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Joint Consensus Initiative. Fasting is not routinely required for determination of a lipid profile: clinical and laboratory implications including flagging at desirable concentration cutpoints-a joint consensus statement from the European Atherosclerosis Society and European Federation of Clinical Chemistry and Laboratory Medicine.
Clin Chem. 2016;62(7):930-946.
PubMedGoogle ScholarCrossref 25.Clarke
R, Shipley
M, Lewington
S,
et al. Underestimation of risk associations due to regression dilution in long-term follow-up of prospective studies.
Am J Epidemiol. 1999;150(4):341-353.
PubMedGoogle ScholarCrossref 27.Knuiman
MW, Divitini
ML, Buzas
JS, Fitzgerald
PE. Adjustment for regression dilution in epidemiological regression analyses.
Ann Epidemiol. 1998;8(1):56-63.
PubMedGoogle ScholarCrossref 28.Jørgensen
AB, Frikke-Schmidt
R, Nordestgaard
BG, Tybjærg-Hansen
A. Loss-of-function mutations in APOC3 and risk of ischemic vascular disease.
N Engl J Med. 2014;371(1):32-41.
PubMedGoogle ScholarCrossref 29.Havel
RJ. Pathogenesis, differentiation and management of hypertriglyceridemia.
Adv Intern Med. 1969;15:117-154.
PubMedGoogle Scholar 30.Saharia
P, Margolis
S, Zuidema
GD, Cameron
JL. Acute pancreatitis with hyperlipemia: studies with an isolated perfused canine pancreas.
Surgery. 1977;82(1):60-67.
PubMedGoogle Scholar 31.Ewald
N, Hardt
PD, Kloer
HU. Severe hypertriglyceridemia and pancreatitis: presentation and management.
Curr Opin Lipidol. 2009;20(6):497-504.
PubMedGoogle ScholarCrossref 32.Czakó
L, Szabolcs
A, Vajda
A,
et al. Hyperlipidemia induced by a cholesterol-rich diet aggravates necrotizing pancreatitis in rats.
Eur J Pharmacol. 2007;572(1):74-81.
PubMedGoogle ScholarCrossref 33.Varbo
A, Benn
M, Tybjærg-Hansen
A, Nordestgaard
BG. Elevated remnant cholesterol causes both low-grade inflammation and ischemic heart disease, whereas elevated low-density lipoprotein cholesterol causes ischemic heart disease without inflammation.
Circulation. 2013;128(12):1298-1309.
PubMedGoogle ScholarCrossref 34.Brunzell
JD, Deeb
SS. Familial lipoprotein Lipase Deficiency, Apo C-II Defiency, and Hepatic Lipase Deficiency. In: Scriver
CR, Beaudet
AL, Sly
WS, Valle
D, eds. The Metabolic & Molecular Bases of Inherited Disease. 8th ed. New York, NY: McGraw-Hill; 2001:2789-2816.
35.Nordstoga
K, Sørby
R, Olivecrona
G, Smith
AJ, Christophersen
B. Pancreatitis in hyperlipemic mink (Mustela vison).
Vet Pathol. 2012;49(3):557-561.
PubMedGoogle ScholarCrossref 36.Nordestgaard
BG. Triglyceride-Rich Lipoproteins and Atherosclerotic Cardiovascular Disease: New Insights From Epidemiology, Genetics, and Biology.
Circ Res. 2016;118(4):547-563.
PubMedGoogle ScholarCrossref 37.Nordestgaard
BG, Zilversmit
DB. Large lipoproteins are excluded from the arterial wall in diabetic cholesterol-fed rabbits.
J Lipid Res. 1988;29(11):1491-1500.
PubMedGoogle Scholar 38.Preiss
D, Tikkanen
MJ, Welsh
P,
et al. Lipid-modifying therapies and risk of pancreatitis: a meta-analysis.
JAMA. 2012;308(8):804-811.
PubMedGoogle ScholarCrossref 40.Garg
R, Chen
W, Pendergrass
M. Acute pancreatitis in type 2 diabetes treated with exenatide or sitagliptin: a retrospective observational pharmacy claims analysis.
Diabetes Care. 2010;33(11):2349-2354.
PubMedGoogle ScholarCrossref 41.Yang
L, He
Z, Tang
X, Liu
J. Type 2 diabetes mellitus and the risk of acute pancreatitis: a meta-analysis.
Eur J Gastroenterol Hepatol. 2013;25(2):225-231.
PubMedGoogle ScholarCrossref 42.Lindkvist
B, Appelros
S, Regnér
S, Manjer
J. A prospective cohort study on risk of acute pancreatitis related to serum triglycerides, cholesterol and fasting glucose.
Pancreatology. 2012;12(4):317-324.
PubMedGoogle ScholarCrossref 44.Christian
JB, Arondekar
B, Buysman
EK, Jacobson
TA, Snipes
RG, Horwitz
RI. Determining triglyceride reductions needed for clinical impact in severe hypertriglyceridemia.
Am J Med. 2014;127(1):36-44.e1.
PubMedGoogle ScholarCrossref 45.Steinberg
WM, Nauck
MA, Zinman
B,
et al; LEADER Trial investigators. LEADER 3—lipase and amylase activity in subjects with type 2 diabetes: baseline data from over 9000 subjects in the LEADER Trial.
Pancreas. 2014;43(8):1223-1231.
PubMedGoogle ScholarCrossref 47.Lippi
G, Valentino
M, Cervellin
G. Laboratory diagnosis of acute pancreatitis: in search of the Holy Grail.
Crit Rev Clin Lab Sci. 2012;49(1):18-31.
PubMedGoogle ScholarCrossref 48.Floyd
A, Pedersen
L, Nielsen
GL, Thorladcius-Ussing
O, Sorensen
HT. Secular trends in incidence and 30-day case fatality of acute pancreatitis in North Jutland County, Denmark: a register-based study from 1981-2000.
Scand J Gastroenterol. 2002;37(12):1461-1465.
PubMedGoogle ScholarCrossref