The association between parenteral nutrition and pancreatic injury in adult patients: a retrospective observational study

BACKGROUND AND OBJECTIVE: Patients on parenteral nutrition (PN) are at high risk of both liver and pancreatic injury. More efforts were focused on liver, however, limited data is available to evaluate the effects of PN on pancreas. Thus, we performed a retrospective observational study to evaluate the association between PN and pancreatic injury in Chinese adult patients. METHODS: Adult patients (18–80 years), who received PN for a week or longer, and with repeated measurements of pancreatic enzymes, were included in the analysis. Pancreatic injury was confirmed by serum level of pancreatic amylase (P-AMYwas 53 U/L or higher) or lipase (LP was 63 U/L or higher), which were evaluated at baseline and following every week during PN duration. Age, sex, body weight, height, diagnosis of diseases, history of diseases, surgery, white blood cell, c-reactive protein, liver and renal function, fasting blood glucose, lipid profile, and daily energy supplied by PN and enteral nutrition were abstracted from medical records. RESULTS: A total number of 190 adult patients (125 men, 65 women) were included in the study. The average age and BMI were 61.8 ± 13.0 years and 21.7±3.3 kg/m(2), while medium serum level of P-AMY and LP were 29.0 U/L (quartile range: 18.0, 47.0) and 33.0 U/L (quartile range: 19.0, 58.0), respectively at baseline. The median duration of PN was 15 days (quartile range: 11.0, 21.0). The prevalence of pancreatic injury was 42.1% (80/190) while it was 28.4% (54/190), 43.3% (77/178), 47.8% (44/92) after one-, two-, and three-week or longer PN adminstration. The proportion of daily energy supplement by PN (OR = 3.77, 95%CI: 1.87, 7.61) and history of infection were positively (OR = 3.00, 95%CI: 1.23, 7.36), while disease history for diabetes mellitus (OR = 0.38, 95%CI: 0.15, 0.98) and cancer (OR = 0.46, 95%CI: 0.23, 0.95), were negetively associated with pancreatic injury. Total bile acids were associated with the increment of P-AMY (beta = 0.98, 95%CI: 0.39, 1.56) and LP (beta = 2.55, 95%CI: 0.98, 4.12) by multi-variate linear regression. CONCLUSION: PN was associated with pancreatic injury, as demonstrated by the increase of both serum P-AMY and LP. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12986-022-00706-z.