A
A prospective cohort study of dietary intake and clinical-nutritional biomarkers in patients with periampullary malignancies undergoing pancreatoduodenectomy undergoing pancreatoduodenectomy in Southeast BrazilA
A
MarcoAurélioRibeiro1✉Emailmarcoas@hcrp.usp.br
Aurélio1
DosSantos1
JoséSebastião2
Perdoná1
Gleici
da
Silva3AndersonNavarro1
Marliere1
1Division of Nutrition and Metabolism, Department of Health Sciences, Ribeirão Preto School of MedicineUniversity of São Paulo (FMRP-USP)Ribeirão PretoBrazil
2Division of Digestive Surgery, Department of Surgery and Anatomy, Ribeirão Preto School of MedicineUniversity of São Paulo (FMRP-USP)Ribeirão PretoBrazil
3Division of Epidemiology and Biostatistics, Department of Social Medicine, Ribeirão Preto School of MedicineUniversity of São Paulo (FMRP-USP)Ribeirão PretoBrazil
Ribeiro, Marco Aurélio1; Dos Santos, José Sebastião2; Perdoná, Gleici da Silva3; Navarro, Anderson Marliere1
1. Division of Nutrition and Metabolism, Department of Health Sciences, Ribeirão Preto School of Medicine, University of São Paulo (FMRP-USP), Ribeirão Preto, Brazil.
2. Division of Digestive Surgery, Department of Surgery and Anatomy, Ribeirão Preto School of Medicine, University of São Paulo (FMRP-USP), Ribeirão Preto, Brazil.
3. Division of Epidemiology and Biostatistics, Department of Social Medicine, Ribeirão Preto School of Medicine, University of São Paulo (FMRP-USP), Ribeirão Preto, Brazil.
*Corresponding author: Marco Aurélio Ribeiro, E-mail: marcoas@hcrp.usp.br
A prospective cohort study of dietary intake and clinical-nutritional biomarkers in patients with periampullary malignancies undergoing pancreatoduodenectomy undergoing pancreatoduodenectomy in Southeast Brazil
A
Abstract
Background
Patients undergoing pancreatoduodenectomy for periampullary malignant neoplasms face a high nutritional risk due to both the underlying disease and the post-operative anatomical and functional changes. Assessing nutritional status in this context is challenging, as traditional biomarkers are influenced by systemic inflammation, and food intake can be compromised by prolonged gastrointestinal symptoms. Studies that integrate clinical, biochemical, and dietary data in a longitudinal approach are still scarce.
Aim
A
To describe and compare the evolution of food intake and clinical-laboratory biomarkers, including nutritional, metabolic, inflammatory, hematological, vitamin, and mineral parameters, in patients who underwent pancreatoduodenectomy for the treatment of periampullary malignant neoplasms. Evaluations were conducted in the pre-operative period and at three and six months after hospital discharge, in relation to a control group.Methods
A
This is a prospective, longitudinal, and controlled cohort study conducted at the Hospital das Clínicas of Ribeirão Preto Medical School, University of São Paulo ((HCFMRP-USP) between July 2023 and December 2024 and the study was approved by the Research Ethics Committee (CAAE: 64695422.6.0000.5440 and 63993522.3.0000.5440), and was registered at ClinicalTrials ID: NCT07159672 (https://clinicaltrials.gov/study/NCT07159672). The study included 24 patients who underwent pancreatoduodenectomy and 20 controls matched by age and gender. Food intake was assessed using a quantitatively adapted Food Frequency Questionnaire, and clinical-nutritional biomarkers (total proteins, albumin, C-reactive protein, hemoglobin, ferritin, vitamins, and minerals) were evaluated at three distinct time points in the surgical group (pre-operative, and at 3 and 6 months after discharge). The control group was evaluated at a single time point. Statistical analyses included longitudinal models and adjusted regressions.Results
Energy intake significantly decreased after surgery (pre-operative: 3444 ± 1130 kcal; 3 months: 1988 ± 826 kcal; p < 0.001), with no spontaneous recovery at 6 months. There was a sustained drop in the intake of essential macro and micronutrients, and serum levels of vitamin E remained reduced during follow-up (p < 0.001). Although not all associations between intake and biomarkers reached statistical significance, clinically relevant correlations were highlighted between iron and hematological parameters, as well as between vitamin E and its serum concentration.
Conclusion
Nutritional assessment of patients undergoing pancreatoduodenectomy for periampullary malignant neoplasms requires the integration of clinical and nutritional data. The multiparametric longitudinal approach proposed in this study applies to different clinical contexts and can contribute to the planning of personalized nutritional strategies in complex surgical-oncological scenarios.
Keywords:
Pancreatoduodenectomy
Nutritional status
Serum biomarkers
Food intake
Periampullary neoplasms
A
A
A
1. Introdução
The pancreas is crucial for metabolic and digestive regulation, producing essential enzymes and hormones for homeostasis and nutrient absorption [1, 2]. Periampullary malignant neoplasms, particularly pancreatic ductal adenocarcinoma, have a high incidence and mortality, with increasing worldwide trends. Late diagnosis, associated with nonspecific symptoms like weight loss and nutritional decline, contributes to a poor prognosis.
Pancreatoduodenectomy (PD), or the Whipple procedure, is the main curative surgical intervention for resectable tumors in the periampullary region [5, 6]. However, PD entails significant post-operative complications, including perioperative mortality of up to 5% and a variable incidence of pancreatic fistula (11.4% to 64.3%). These complications highlight the importance of assessing nutritional status and food intake after surgery, as malnutrition worsens recovery and impacts clinical outcomes.
Patients undergoing PD frequently present with nutritional challenges, stemming from the loss of pancreatic tissue, exocrine insufficiency, malabsorption, and protein-calorie malnutrition [8, 9]. These disorders result from multiple factors, including reduced intake, digestive changes, chronic inflammation, and cachexia, which hinder rehabilitation and prognosis [10, 11, 12]. Optimizing perioperative nutrition can significantly improve the recovery and clinical outcomes of these patients.
The prevalence of malnutrition in pancreatic cancer patients is high, and its early identification and management are critical for therapeutic success [14, 15]. Pre-operative nutritional interventions contribute to reducing post-operative complications [6, 16].
Food intake after PD is often inadequate, exacerbating nutritional status. The post-operative period is marked by difficulties in meeting daily energy and protein needs due to surgical stress and gastrointestinal changes [18, 10].
A
A
Dietary management requires a gradual progression from liquids to solids and rigorous monitoring of food intake and clinical biomarkers, as guided by international guidelines.Clinical and nutritional biomarkers have gained prominence in the evaluation of nutritional risk and status, showing predictive value for post-operative complications. Laboratory parameters, such as total proteins, albumin, hemoglobin, lymphocytes, transferrin, ferritin, fat-soluble vitamins, trace elements, and inflammatory markers, are widely used for monitoring and individualizing care [20, 21, 22]. Composite indices have shown a correlation with complications, length of hospital stay, survival, and quality of life after PD [23, 24, 25]. The detailed measurement of these markers allows for the early identification of deficiencies and the adaptation of nutritional strategies.
Although there have been advances in nutritional assessment and the use of biomarkers, no prospective and longitudinal studies have been found that, in an integrated manner, describe dietary patterns, the evolution of laboratory parameters, and their functional implications in the post-operative period of pancreatoduodenectomy, especially within the Brazilian context. Considering that these patients exhibit a significant decline in food intake and adverse changes in clinical-nutritional indicators, with a negative impact on recovery and clinical outcomes, it is necessary to systematically investigate these aspects to inform more precise and effective nutritional strategies.
2. Objective
The objective of the present study was to prospectively and longitudinally evaluate the evolution of food intake and clinical-laboratory biomarkers, including nutritional, metabolic, inflammatory, hematological, vitamin, and mineral parameters, in patients who underwent pancreatoduodenectomy for the treatment of periampullary malignant neoplasms. Patients were assessed during the pre-operative period and at three and six months after hospital discharge, with comparisons to a control group. The study also sought to investigate the correlations between the intake of protein, iron, zinc, vitamin A, vitamin E, folic acid, and vitamin B12 and their respective serum biomarkers throughout the follow-up, aiming to support individualized nutritional strategies and optimize multidisciplinary care.
3. Materials and Methods
3.1 Ethical Considerations
The study was approved by the Research Ethics Committee of the Ribeirão Preto Medical School, University of São Paulo (CAAE: 64695422.6.0000.5440 and 63993522.3.0000.5440), and the study was registered at ClinicalTrials ID: NCT07159672 (https://clinicaltrials.gov/study/NCT07159672). The authors report no conflicts of interest.
A
A
All participants signed an informed consent form, in accordance with the Declaration of Helsinki. Data privacy was ensured through individual coding.3.2 Experimental Design
This was a prospective, longitudinal, and controlled cohort study conducted at the Hospital das Clínicas of the Faculty of Medicine of Ribeirão Preto at the University of São Paulo (HCFMRP-USP), a tertiary referral center for the treatment of periampullary malignant neoplasms and highly complex oncological surgeries in Southeast Brazil. Participants were recruited consecutively between July 2023 and December 2024.
The surgical group, consisting of patients who underwent pancreatoduodenectomy for the management of these conditions, was evaluated pre-operatively and at three and six months after hospital discharge.
The control group, matched by age and sex, was evaluated at a single time point. The service has a specialized multidisciplinary team in clinical nutrition and metabolic support, laboratory infrastructure, and standardized care protocols. In each phase, clinical data were collected, food intake was assessed, and laboratory samples were collected under uniform conditions by a trained team. The analyzed data are part of the information database originating from the main author's doctoral thesis, which is used as a basis for different publications with distinct objectives and analytical perspectives.
3.3 Participants
The sample size was calculated for a statistical power of 80% and a significance level of 5% (α = 0.05). In the pancreatoduodenectomy (PD) group, 20% was added to account for anticipated losses due to death in this population.
The experimental group consisted of 24 adults and elderly individuals (18–80 years) who underwent PD for the treatment of periampullary malignant neoplasms. They were evaluated pre-operatively, and at three and six months after hospital discharge. All surgeries were performed by the same team, following standardized surgical techniques.
A
The choice between the pylorus-preserving or the classic Whipple technique followed the institutional routine, with pancreatic reconstruction via a terminolateral duct-to-mucosa pancreaticojejunostomy.The control group was formed by 20 individuals matched for age and sex, selected from patients who underwent upper digestive endoscopy with normal findings or nonspecific alterations, including functional dyspepsia. The selection process involved a review of medical records to exclude chronic metabolic or gastrointestinal diseases or other conditions that could interfere with the variables studied.
Individuals who had surgery at other facilities, had tumor recurrence or metastases at the start of the study, or had clinical conditions that could interfere with the analyzed outcomes were excluded. To ensure uniformity and minimize variations, all clinical, dietary, and laboratory assessments were conducted by the same, previously trained team, using standardized protocols and instruments. The recruitment, inclusion, follow-up, and final sample analysis process is detailed in the flowchart presented in Fig. 1. All losses during follow-up, as well as the reasons for exclusion, were documented, ensuring transparency in the design and allowing for the reproducibility of the study steps.
Fig. 1
Flow of study volunteers, evaluating patients who underwent pancreatoduodenectomy (PD) and individuals in the control group. The flowchart presents the initial number of recruited volunteers, the inclusion and exclusion criteria applied to each group, the losses during follow-up, and the final number of participants included in the analyses.
3.4 Study Variables
The outcome variables were food intake, including macronutrients and micronutrients, and clinical-nutritional biomarkers. The exposure variables were the study group (pancreatoduodenectomy vs. control) and the time of evaluation (pre-operative, three, and six months after hospital discharge). No relevant confounding factors were identified other than age and sex, which were controlled by matching. The inflammatory status was monitored by C-reactive protein (CRP) and ferritin, considered both a marker of iron stores and an acute-phase protein.
A
The diagnostic criteria and cut-off points for classifying the laboratory variables followed institutional reference values, and the food composition was obtained from specific tables for macro and micronutrients.3.5 Assessment of Food Intake
Food intake was assessed using the Food Frequency Questionnaire (FFQ) developed and validated in the Brazilian Longitudinal Study on Adult Health (ELSA-Brasil), originally composed of 114 food items. This instrument was selected because it has demonstrated validity and reproducibility in the Brazilian adult population and allows for a more comprehensive estimation of habitual consumption than 24-hour dietary recalls, thus minimizing recall bias. In the present study, the FFQ was adapted for quantitative application, fully retaining its original list of foods but detailing the portion size, frequency of consumption, and preparation method for each item. The frequencies were converted to average daily consumption, with "times per week" divided by seven and "times per month" divided by thirty. The resulting quantity (in grams, milligrams, micrograms, or kilocalories) was calculated based on national and international food composition tables, processed using the Nutrium® software.
A
This adaptation made it possible to more accurately estimate the average intake of macro- and micronutrients over 30 days, more closely approximating the participants' actual consumption and allowing for longitudinal comparisons between evaluation times and between groups.3.6 Clinical-Nutritional Biomarkers
Laboratory tests were part of the routine care of the Gastrosurgery service and were processed in the HCFMRP-USP laboratories according to institutional protocols. Serum total proteins (6.0–8.3 g/dL) and albumin (3.5–5.0 g/dL) were determined by a colorimetric method, while C-reactive protein (CRP), an inflammatory marker, was evaluated by immunoturbidimetry (reference < 3 mg/L). A complete blood count was obtained using an automated analyzer (ABX Pentra DX120 – Horiba Medical), considering hemoglobin values of 13.0–17.5 g/dL for men and 12.0–15.5 g/dL for women, and an absolute lymphocyte count between 1.0 and 4.0 × 10³/mm³.
Iron metabolism was investigated by measuring serum iron (65–175 µg/dL in men and 50–170 µg/dL in women), ferritin (30–300 ng/mL in men and 15–150 ng/mL in women), and latent iron-binding capacity (240–450 µg/dL), all using colorimetric or chemiluminescence methods, with transferrin calculated according to Vannucchi et al. (200–360 mg/dL) and transferrin saturation (20–50%). The hepatic enzymes Aspartate aminotransferase - AST (10–40 U/L) and Alanine aminotransferase - ALT (7–56 U/L) were measured by an optimized UV method, while renal function was assessed by urea (10–40 mg/dL) and creatinine (0.7–1.3 mg/dL in men and 0.6–1.1 mg/dL in women) using a kinetic method.
Vitamins B9 (folic acid; 3.0–17.0 ng/mL) and B12 (200–900 pg/mL) were quantified by chemiluminescence, and vitamins A (retinol; 30–80 µg/dL) and E (α-tocopherol; 0.5–2.0 mg/dL) by high-performance liquid chromatography. Minerals such as copper (70–140 µg/dL in men and 80–155 µg/dL in women) and zinc (60–120 µg/dL) were analyzed by atomic absorption spectrophotometry. The presence of steatorrhea was investigated by the qualitative Sudan III test, with a negative result considered normal. All reference values used correspond to the parameters adopted by HCFMRP-USP.
3.7 Bias Control
Several measures were adopted to minimize the risk of bias. Selection bias was reduced by consecutively recruiting all eligible patients during the study period, with no refusals in the surgical group, and by selecting the control group matched for age and sex from patients who underwent upper digestive endoscopy with normal findings or nonspecific alterations, following a rigorous screening of medical records to exclude chronic metabolic or gastrointestinal diseases. Information bias was minimized by standardizing the procedures for collecting clinical, dietary, and laboratory data, which were carried out by the same previously trained team using validated instruments. Measurement bias was prevented by using laboratory methods with internal quality control and institutional reference values, as well as by applying the adapted Food Frequency Questionnaire from a nationally validated instrument, ensuring temporal and inter-group comparability. Intra- and inter-observer variability was controlled by maintaining the same trained team throughout all stages of the study, using standardized operational protocols for measurements and interviews.
3.8 Statistical Analysis
Analyses were conducted in RStudio (version 2024.12.1 + 563). The normality of continuous variables was checked using the Shapiro-Wilk and Kolmogorov-Smirnov tests. Variables with a normal distribution were expressed as mean ± standard deviation, and skewed variables as median and interquartile range. Categorical variables were presented as absolute and relative frequencies, without performing inferential tests, as they were used exclusively for sample characterization. These variables were coded in the database in a binary or ordinal manner as applicable. The adopted significance level was 5%, with a Holm-Bonferroni correction for multiple comparisons.
3.8.1 Treatment of Missing Data
Considering the longitudinal design and the follow-up loss rate (~ 33%), multiple imputation by chained equations (MICE) was applied, generating 50 imputed sets in 20 iterations [28, 29]. The final estimates were combined according to Rubin’s Rules, in which the total variance (T) incorporates both the within-imputation variance (U) and the between-imputation variance (B), as per T = U + (1 + 1/m) B, where m is the number of imputed sets.
3.8.2 Intragroup Comparisons
The temporal evolution of variables was evaluated using Mixed-Effect Linear Models (MLM) for approximately normal outcomes, with fixed effects for time, group, and time × group interaction, and a random intercept per participant. Parameters were estimated by restricted maximum likelihood, and the residual covariance structure was defined based on the Akaike and Bayes criteria, with inspection of the residuals. For skewed variables, Generalized Mixed-Effect Linear Models (GLMM) were used, with a Gamma family and log link function, maintaining the same specification of fixed and random effects [32, 33]. Comparisons between time points (pre-operative vs. 3 months; 3 months vs. 6 months; pre-operative vs. 6 months) were based on estimated marginal means, with p-value adjustment by the Holm-Bonferroni method.
3.8.3 Intergroup Comparisons
Differences between groups at each time point were analyzed by tests appropriate for the distribution and homogeneity of variances (Levene's test). The Student’s t-test was used when both conditions were met, Welch’s t-test when variances were unequal, and the non-parametric Mann-Whitney or Wilcoxon tests when normality was not adhered to. The test selection was automated by a script in R, ensuring traceability and standardization.
3.8.4 Association Models
Associations between nutrient intake and corresponding biomarkers were evaluated by multiple linear regression models, adjusted for C-reactive protein, ferritin, age, and sex, defined by clinical relevance and prior evidence. The previously described criteria for significance and correction for multiple comparisons were maintained.
3.8.5 Analytical Integrity
All statistical analyses were independently reviewed by two researchers to ensure accuracy and reproducibility. The same raw datasets were reanalyzed for numerical and interpretative consistency. The study complied with the STROBE Statement, ensuring methodological rigor and transparency in reporting.
4. Results
4.1 Clinical characteristics of the sample
Table 1 presents the characteristics of the participants. The sample included 24 patients who underwent pancreatoduodenectomy and 20 controls, with the groups being comparable in age and sex distribution. Pancreatic head adenocarcinoma was the predominant diagnosis, followed by adenocarcinoma of the ampulla of Vater. The TNM staging analysis revealed a predominance of locally advanced tumors, especially in the pT2 stage. The classification followed the criteria of the 8th edition of the UICC/AJCC [34].
Variable | PD Group (n = 24) | Control Group (n = 20) |
|---|---|---|
Age (years) | 62,9 ± 11,9 | 61,8 ± 12,5 |
Male sex | 15 (62,5%) | 12 (60%) |
Famale sex | 9 (37,5%) | 8 (40%) |
Histopathological diagnosis of the PD Group | (n,%) | |
Pancreatic head adenocarcinoma | 11 (45,8%) | |
Ampullary adenocarcinoma | 8 (33,3%) | |
Uncinate process adenocarcinoma | 2 (8,3%) | |
Pancreatic ductal adenocarcinoma | 1 (4,2%) | |
Biliary carcinoma | 2 (8,3%) | |
TNM classification of the PD Group | (n,%) | |
CN1 M0 G2 | 1 (6,20) | |
pT1 N0 M0 | 1 (6,20) | |
pT1 PNI1 M0 | 1 (6,20) | |
pT1C N0 M0 | 1 (6,20) | |
pT2 N0 M0 | 3 (18,80) | |
pT2 N2 M0 | 2 (12,50) | |
pT2 N1 M0 | 2 (12,50) | |
pT3 N1 M0 | 1 (6,20) | |
pT3B N0 M0 | 1 (6,20) | |
pT3B N1 M0 | 1 (6,20) | |
pT4 N0 M0 | 1 (6,20) | |
TNM: Tumor staging system of the Union for International Cancer Control (UICC) and the American Joint Committee on Cancer (AJCC), 8th edition (2017) [34], based on the extent of the primary tumor (T), lymph node involvement (N), and presence of distant metastases (M). | ||
4.2 Dietary intake
The Table 2 presents the evolution of dietary intake in patients undergoing pancreaticoduodenectomy, with assessments at the preoperative period, and at 3 and 6 months after hospital discharge. The results demonstrated a significant and sustained reduction in total energy intake after surgery, with no spontaneous recovery during the follow-up period. This pattern was consistently observed for all macronutrients evaluated (carbohydrates, lipids, and proteins), as well as for fiber intake. Among micronutrients, there was a generalized decline in the intake of essential minerals (calcium, iron, potassium, selenium, and zinc) and vitamins (A, C, and E), with persistent depletion throughout the 6-month follow-up. These findings highlight ongoing nutritional impairment in the postoperative period, indicating the need for targeted nutritional intervention to prevent long-term deficiencies.
Preoperative (n = 24) | 3 months after discharge (n = 18) | PD – 6 months after discharge (n = 16) | ||||
|---|---|---|---|---|---|---|
Mean ± SD or Median (P25–P75) | p1 | Mean ± SD or Median (P25–P75) | p2 | Mean ± SD or Median (P25–P75) | p3 | |
Energy (kcal) | 3511.64 (2586.19–4380.52) | z = 6.080; p = 0.000 | 1730.12 (1338.49–2359.43) | z = − 0.320; p = 0.945 | 1650.53 (1412.58–2283.87) | z = 6.426; p = 0.000 |
Carbohydrate (g) | 491.90 (364.31–582.55) | z = 6.545; p = 0.000 | 276.81 (208.92–305.89) | z = − 0.954; p = 0.606 | 287.95 (235.57–335.08) | z = 7.501; p = 0.000 |
Lipid (g) | 101.17 (79.19–127.72) | z = 5.381; p = 0.000 | 50.79 (40.26–67.01) | z = 0.199; p = 0.978 | 44.15 (40.12–70.76) | z = 5.188; p = 0.000 |
Protein (g) | 142.28 (114.92–215.60) | z = 6.399; p = 0.000 | 92.44 (71.28–102.57) | z = 0.191; p = 0.980 | 84.26 (65.27–115.16) | z = 6.230; p = 0.000 |
Fiber (g) | 57.23 (45.48–73.02) | z = 6.510; p = 0.000 | 33.17 (29.95–42.07) | z = − 0.168; p = 0.985 | 34.89 (27.88–44.66) | z = 6.675; p = 0.000 |
Calcium (mg) | 1831.68 ± 490.50 | t(46) = 5.921; p = 0.000 | 1249.66 ± 464.36 | t(46) = 0.288; p = 0.955 | 1219.86 ± 441.27 | t(46) = 5.632; p = 0.000 |
Iron (mg) | 33.34 (27.75–44.23) | z = 4.942; p = 0.000 | 21.35 (15.86–25.65) | z = -0.038; p = 0.999 | 22.01 (16.75–24.20) | z = 4.959; p = 0.000 |
Potassium (mg) | 8486.07 (6541.50–10213.13) | z = 5.406; p = 0.000 | 5339.72 (4495.65–6299.40) | z = 0.485; p = 0.878 | 5194.51 (4453.46–6060.92) | z = 4.938; p = 0.000 |
Selenium (µg) | 74.04 ± 34.12 | t(46) = 5.062; p = 0.000 | 46.04 ± 21.55 | t(46) = 0.357; p = 0.932 | 43.92 ± 20.31 | t(46) = 4.705; p = 0.000 |
Zinc (mg) | 21.82 (16.48–27.97) | z = 5.294; p = 0.000 | 12.76 (10.10–15.11) | z = − 0.729; p = 0.746 | 13.46 (10.42–15.91) | z = 6.036; p = 0.000 |
Sodium (mg) | 1990.23 (1606.37–2825.68) | z = 6.765; p = 0.000 | 1394.30 (928.46–1671.63) | z = 1.285; p = 0.403 | 1136.58 (758.85–1540.41) | z = 5.491; p = 0.000 |
Vitamin A (mg) | 5.71 (3.91–10.41) | z = 3.221; p = 0.007 | 4.13 (3.25–6.72) | z = − 0.212; p = 0.976 | 4.26 (3.03–6.70) | z = 3.446; p = 0.005 |
Vitamin C (mg) | 412.88 (255.65–696.35) | z = 4.239; p = 0.000 | 321.84 (230.29–431.98) | z = 1.133; p = 0.493 | 243.90 (196.64–426.20) | z = 3.102; p = 0.011 |
Vitamin E (mg) | 6.38 (3.25–7.66) | z = 4.806; p = 0.000 | 3.50 (2.97–5.19) | z = 0.194; p = 0.980 | 3.71 (2.79–5.33) | z = 4.641; p = 0.000 |
kcal: kilocalories; g: grams; mg: milligrams; mcg: micrograms. Data are presented as mean ± standard deviation (SD) or as median (P25–P75), according to the statistical distribution observed at each time point. p-values refer to the statistical models applied for comparisons between time points: p¹ = preoperative vs. 3 months; p² = 3 months vs. 6 months; p³ = preoperative vs. 6 months. Linear mixed models (t-test, with degrees of freedom estimated by the Satterthwaite method) were used for variables with approximately normal distribution, and generalized linear models (z-statistic) for variables with non-normal distribution, according to the adequacy of each variable. Holm–Bonferroni correction was applied for multiple comparisons. p-values ≤ 0.05 were considered statistically significant. | ||||||
A
Table 3 presents the dietary intake data at the preoperative period, and at 3 and 6 months after hospital discharge of patients undergoing pancreaticoduodenectomy, compared with the control group.At baseline, the surgical group showed higher intake of energy (p = 0.010) and carbohydrates (p = 0.001), as well as significantly higher values for fiber, calcium, iron, potassium, zinc, sodium, vitamin A, and vitamin C. Conversely, selenium and vitamin E intake were lower (both p < 0.001), with no relevant differences for the other nutrients evaluated.
At three months after discharge, it became evident that the operated patients presented significantly lower intake of energy (p = 0.002), protein (p = 0.002), lipids (p = 0.003), and carbohydrates (p = 0.006) compared with controls, but higher intake of fiber (p = 0.023), calcium (p = 0.001), iron (p = 0.001), potassium (p < 0.001), sodium (p = 0.024), and vitamin A (p < 0.001), with persistent deficiencies in selenium and vitamin E (both p < 0.001).
At six months after surgery, the pancreaticoduodenectomy group maintained a lower intake of energy (p = 0.002), carbohydrates (p = 0.035), lipids (p = 0.002), and protein (p = 0.001), along with higher intake of fiber (p = 0.037), calcium (p = 0.001), iron (p < 0.001), potassium (p < 0.001), and vitamin A (p < 0.001), while inadequacies in selenium and vitamin E persisted (both p < 0.001). The other nutrients did not show statistically significant differences at any of the evaluated time points.
4.3 Clinical–Nutritional Laboratory Analyses
Table 4 presents the laboratory results of patients who underwent pancreatoduodenectomy in the pre-operative period, and at three and six months after hospital discharge. The concentrations of total proteins and albumin remained stable during follow-up (p > 0.05). C-reactive protein showed a significant reduction between the pre-operative and post-operative periods (p < 0.001).
Tabela 3. Summary measures of dietary intake variables comparing the Control Group and PD patients at baseline, 3 months, and 6 months after hospital discharge. | |||||||||
|---|---|---|---|---|---|---|---|---|---|
Control Group (n = 20) | Preoperative (n = 24) | 3 months after discharge (n = 18) | PD – 6 months after discharge (n = 16) | ||||||
Média ± DP | Mediana (P25-P75%) | Mean ± SD or Median (P25–P75) | p1 | Mean ± SD or Median (P25–P75) | p2 | Mean ± SD or Median (P25–P75) | p3 | ||
Energy (kcal) | 2731.14 ± 549.94 | 2591.00 (2296.50–3083.14) | 3444.23 ± 1130.73 | t(34.5) = 2.727; p = 0.010 | 1730.12 (1338.49–2359.43) | U = 105; p = 0.002 | 1650.53 (1412.58–2283.87) | U = 108; p = 0.002 | |
Carbohydrate (g) | 347.36 ± 75.09 | 334.73 (289.48–390.41) | 481.90 ± 166.04 | t(33.3) = 3.557; p = 0.001 | 276.81 (208.92–305.89) | U = 124; p = 0.006 | 287.95 (235.57–335.08) | U = 150; p = 0.035 | |
Lipid (g) | 84.27 ± 25.76 | 81.70 (67.36–100.41) | 105.38 ± 45.81 | t(37.3) = 1.922; p = 0.062 | 50.79 (40.26–67.01) | U = 113; p = 0.003 | 44.15 (40.12–70.76) | U = 109; p = 0.002 | |
Protein (g) | 134.85 ± 43.32 | 138.20 (100.65–157.95) | 159.55 ± 62.06 | t(42) = 1.500; p = 0.141 | 92.44 (71.28–102.57) | U = 110; p = 0.002 | 84.26 (65.27–115.16) | U = 99; p = 0.001 | |
Fiber (g) | 28.65 ± 9.90 | 25.80 (22.58–32.31) | 57.23 (45.48–73.02) | U = 446; p = 0.000 | 33.17 (29.95–42.07) | U = 337; p = 0.023 | 34.89 (27.88–44.66) | U = 329; p = 0.037 | |
Calcium (mg) | 742.81 ± 403.79 | 525.60 (445.52–1062.96) | 1875.57 (1509.57–2216.93) | U = 456; p = 0.000 | 1146.75 (874.98–1569.06) | U = 387; p = 0.001 | 1136.41 (878.55–1579.53) | U = 384; p = 0.001 | |
Iron (mg) | 14.65 ± 3.85 | 14.14 (12.40–17.92) | 33.98 ± 10.72 | t(29.8) = 8.228; p = 0.000 | 21.35 (15.86–25.65) | U = 384; p = 0.001 | 22.01 (16.75–24.20) | U = 392; p = 0.000 | |
Potassium (mg) | 3506.64 ± 955.95 | 3680.39 (2683.05–4257.52) | 8411.52 ± 2490.51 | t(30.7) = 8.894; p = 0.000 | 5339.72 (4495.65–6299.40) | U = 425; p = 0.000 | 5194.51 (4453.46–6060.92) | U = 414; p = 0.000 | |
Selenium (µg) | 210.71 ± 132.09 | 203.05 (67.66–331.17) | 74.04 ± 34.12 | t(21.1) = -4.503; p = 0.000 | 46.04 ± 21.55 | t(19.8) = − 5.514; p = 0.000 | 43.92 ± 20.31 | t(19.7) = -5.592; p = 0.000 | |
Zinc (mg) | 13.48 ± 3.23 | 12.95 (10.78–16.33) | 22.89 ± 8.71 | t(30.2) = 4.907; p = 0.000 | 12.76 (10.10–15.11) | U = 228; p = 0.786 | 13.46 (10.42–15.91) | U = 237; p = 0.953 | |
Sodium (mg) | 1032.30 ± 437.92 | 930.65 (678.40–1362.72) | 1990.23 (1606.37–2825.68) | U = 433; p = 0.000 | 1394.30 (928.46–1671.63) | U = 336; p = 0.024 | 1136.58 (758.85–1540.41) | U = 283; p = 0.316 | |
Vitamin A (mg) | 1.54 ± 0.77 | 1.23 (0.96–1.86) | 5.71 (3.91–10.41) | U = 467; p = 0.000 | 4.13 (3.25–6.72) | U = 459; p = 0.000 | 4.26 (3.03–6.70) | U = 453; p = 0.000 | |
Vitamin C (mg) | 264.34 ± 151.57 | 307.20 (94.04–387.40) | 412.88 (255.65–696.35) | U = 345; p = 0.014 | 321.84 (230.29–431.98) | U = 294; p = 0.207 | 243.90 (196.64–426.20) | U = 266; p = 0.548 | |
Vitamin E (mg) | 12.51 ± 3.89 | 12.95 (9.30–14.05) | 6.38 (3.25–7.66) | U = 41; p = 0.000 | 3.94 ± 1.88 | t(26.3) = -9.011; p = 0.000 | 3.71 (2.79–5.33) | t(26.2) = -9.124; p = 0.000 | |
kcal: kilocalories; g: grams; mg: milligrams; mcg: micrograms. Data are presented as mean ± standard deviation (SD) or as median (P25–P75), according to the statistical distribution observed in each comparison between groups. Comparisons were performed between the Control Group and patients undergoing pancreaticoduodenectomy (PD) at the following time points: p¹ = Control vs. Preoperative; p² = Control vs. 3 months after hospital discharge; p³ = Control vs. 6 months after hospital discharge. Variables with normal distribution and homogeneity of variance were analyzed using Student’s t-test; in cases of heterogeneity of variance, Welch’s t-test was applied, with the respective degrees of freedom indicated in the table. For variables with nonparametric distribution, the Mann–Whitney U test was used. Correction for multiple comparisons was performed using the Holm–Bonferroni method. p-values ≤ 0.05 were considered statistically significant. | |||||||||
Ferritin showed a progressive decrease over time (p = 0.002). Iron metabolism markers, including serum iron, transferrin, transferrin saturation, and latent iron-binding capacity, showed no significant changes between the evaluated periods (p > 0.05).
The hepatic enzymes AST and ALT showed a significant reduction between the pre-operative and post-operative periods (p < 0.001 for both).
As for micronutrients, folic acid levels increased (p = 0.020) and vitamin E levels decreased over time (p = 0.016). Vitamin B12 and zinc levels remained stable (p > 0.05). Copper showed a significant reduction during follow-up (p < 0.001).
The prevalence of steatorrhea progressively increased at six months after hospital discharge. Analysis of the qualitative Sudan test revealed a progressive increase in the prevalence of steatorrhea among patients who underwent pancreatoduodenectomy over time. At three months post-discharge, 27.8% tested positive, distributed among 11.1% with an intensity of +, 11.1% with ++, and 5.6% with +++. At six months, the positivity increased to 43.8%, with 6.3% +, 12.5% ++, and 25.0% +++.
Table 5 presents the comparison of biochemical parameters between pancreatoduodenectomy patients in the pre-operative period, and at 3 and 6 months after discharge, compared with the control group.
Preoperative (n = 24) | 3 months after discharge (n = 18) | PD – 6 months after discharge (n = 16) | ||||
|---|---|---|---|---|---|---|
Mean ± SD or Median (P25–P75) | p1 | Mean ± SD or Median (P25–P75) | Mean ± SD or Median (P25–P75) | p1 | Mean ± SD or Median (P25–P75) | |
Total Proteins (g/dL) | 6.87 (6.74–7.24) | z = 0.807; p = 0.699 | 6.79 (6.44–7.03) | z = 0.122; p = 0.992 | 6.76 (6.36–7.37) | z = 0.929; p = 0.622 |
Albumin (g/dL) | 4.04 (3.45–4.46) | z = -0.646; p = 0.795 | 4.22 (3.92–4.44) | z = 0.037; p = 0.999 | 4.19 (3.85–4.45) | z = -0.609; p = 0.815 |
C-Reactive Protein (mg/L) | 6.19 (2.64–17.03) | z = 5.416; p = 0.000 | 0.77 (0.40–4.23) | z = 2.045; p = 0.102 | 0.40 (0.40–0.62) | z = 6.966; p = 0.000 |
Blood Glucose (mg/dL) | 121.54 (97.36–154.50) | z = 0.088; p = 0.996 | 122.08 (84.74–140.50) | z = 1.374; p = 0.355 | 115.48 (88.98–122.20) | z = 1.457; p = 0.312 |
Hemoglobin (g/dL) | 12.62 ± 2.25 | t(46) = 0.462; p = 0.889 | 12.42 ± 1.57 | t(46) = 0.745; p = 0.738 | 12.10 ± 2.02 | t(46) = 1.207; p = 0.455 |
Lymphocytes (x10³/mm³) | 2233.33 ± 1241.55 | t(46) = 0.199; p = 0.978 | 2170.83 ± 1019.16 | t(46) = -0.726; p = 0.749 | 2398.95 ± 1028.19 | t(46) = -0.527; p = 0.858 |
Serum Iron (µg/dL) | 87.40 ± 38.18 | t(46) = -0.709; p = 0.759 | 95.03 ± 49.70 | t(46) = 0.489; p = 0.877 | 89.77 ± 38.00 | t(46) = -0.22; p = 0.974 |
Ferritin (ng/mL) | 336.00 (93.67–828.80) | z = 1.86; p = 0.015 | 254.60 (159.27–554.30) | z = 1.519; p = 0.282 | 189.60 (119.40–401.50) | z = 3.37; p = 0.002 |
UIBC (µg/dL) | 284.25 (251.53–311.06) | z = -0.711; p = 0.757 | 296.85 (276.77–307.78) | z = -0.215; p = 0.975 | 294.35 (282.52–314.73) | z = -0.926; p = 0.624 |
Transferrin (mg/dL) | 227.40 (201.22–248.85) | z = -0.710; p = 0.755 | 237.48 (221.42–246.22) | z = -0.211; p = 0.970 | 235.48 (226.02–251.78) | z = -0.921; p = 0.610 |
Transferrin Saturation (%) | 31.45 ± 14.67 | t(46) = -0.17; p = 0.984 | 32.07 ± 15.61 | t(46) = 0.491; p = 0.876 | 30.29 ± 13.20 | t(46) = 0.321; p = 0.945 |
Urea (mg/dL) | 34.72 (25.59–41.40) | z = 0.527; p = 0.858 | 31.75 (23.37–38.97) | z = -0.539; p = 0.852 | 31.25 (21.96–40.44) | z = -0.024; p = 1000 |
Creatinine (mg/dL) | 0.92 (0.80–1.16) | z = 1.022; p = 0.563 | 0.84 (0.74–1.14) | z = -0.358; p = 0.932 | 0.87 (0.78–1.09) | z = -1.378; p = 0.352 |
Aspartate Aminotransferase (AST) (U/L) | 147.88 (81.54–267.31) | z = 8.871; p = 0.000 | 29.61 (24.15–38.55) | z = 0.896; p = 0.643 | 27.85 (25.46–38.81) | z = 9.504; p = 0.000 |
Alanine Aminotransferase (ALT) (U/L) | 146.68 (71.69–253.93) | z = 10.462; p = 0.000 | 29.73 (16.52–39.14) | z = -0.504; p = 0.869 | 34.09 (23.50–42.54) | z = 9.957; p = 0.000 |
Folic Acid (ng/mL) | 13.85 (9.91–18.99) | z = -2.09; p = 0.092 | 20.67 (10.81–24.00) | z = -0.592; p = 0.825 | 21.52 (13.80–24.00) | z = -2.684; p = 0.020 |
Vitamin B12 (pg/mL) | 478.00 (410.00–688.00) | z = -1.033; p = 0.556 | 561.50 (382.50–1047.00) | z = 2.017; p = 0.108 | 359.50 (313.00–544.50) | z = 1.001; p = 0.577 |
Vitamin A (µg/dL) | 0.30 (0.19–0.35) | z = -1.161; p = 0.477 | 0.33 (0.27–0.40) | z = -0.217; p = 0.974 | 0.30 (0.26–0.38) | z = -1.378; p = 0.353 |
Vitamin E (mg/dL) | 7.46 (6.11–8.60) | z = 1.949; p = 0.125 | 6.77 (5.93–9.34) | z = 2.767; p = 0.016 | 4.98 (4.88–6.13) | z = 4.723; p = 0.000 |
Copper (µg/dL) | 139.00 (105.35–152.45) | z = 4.729; p = 0.000 | 103.00 (87.35–113.35) | z = 1.179; p = 0.466 | 93.70 (71.80–105.85) | z = 5.900; p = 0.000 |
Zinc (µg/dL) | 87.50 (58.25–96.00) | z = -1.512; p = 0.285 | 77.00 (73.62–86.62) | z = 1.509; p = 0.287 | 80.75 (69.88–89.53) | z = -0.005; p = 1000 |
UIBC: Unsaturated Iron-Binding Capacity; g/dL: grams per deciliter; mg/dL: milligrams per deciliter; mg/L: milligrams per liter; µg/dL: micrograms per deciliter; ng/mL: nanograms per milliliter; pg/mL: picograms per milliliter; x10³/mm³: thousand cells per cubic millimeter; U/L: units per liter; %: percentage. Data are expressed as mean ± SD or median (P25–P75%), according to the statistical distribution observed in each comparison. P-values refer to the statistical models applied for comparisons between time points: p¹ = pre-operative vs. 3 months; p² = 3 months vs. 6 months; p³ = pre-operative vs. 6 months. Mixed-effect linear models (t-test, with degrees of freedom estimated by Satterthwaite's method) were used for variables with an approximately normal distribution, and generalized linear models (z-statistic) for variables with a non-normal distribution, according to the suitability of each variable. The Holm-Bonferroni method was used for the correction of multiple comparisons. P-values ≤ 0.05 were considered statistically significant. | ||||||
Control Group (n = 20) | Preoperative (n = 24) | 3 months after discharge (n = 18) | PD – 6 months after discharge (n = 16) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
Mean ± SD | Mean ± SD or Median (P25–P75) | p1 | Mean ± SD or Median (P25–P75) | Mean ± SD or Median (P25–P75) | p1 | Mean ± SD or Median (P25–P75) | p3 | |||
Total Proteins (g/dL) | 6.99 ± 0.6 | 6.97 (6.7–7.32) | 6.86 (6.74–7.24) | U = 203; p = 0.389 | 6.79 ± 0.46 | t(36) = -1.273; p = 0.210 | 6.77 ± 0.6 | t(34) = -1.211; p = 0.233 | ||
Albumin (g/dL) | 4.46 ± 0.29 | 4.51 (4.39–4.58) | 3.95 ± 0.61 | t(34.2) = -3.583; p = 0.181 | 4.22 (3.92–4.44) | U = 136.5; p = 0.015 | 4.19 (3.85–4.45) | U = 144; p = 0.103 | ||
C-Reactive Protein (mg/L) | 0.52 ± 0.32 | 0.4 (0.4–0.41) | 6.19 (2.64–17.03) | U = 451; p = 0.000 | 0.77 (0.4–4.23) | U = 338.; p = 0.010 | 0.40 (0.4–0.96) | U = 262.5; p = 0.509 | ||
Blood Glucose (mg/dL) | 115.31 ± 68.09 | 96.15 (85.56-109.36) | 121.54 (97.36–154.5) | U = 312; p = 0.092 | 122.08 (84.74–140.5) | U = 291; p = 0.234 | 115.48 (88.98–122.2) | U = 262; p = 0.612 | ||
Hemoglobin (g/dL) | 13.74 ± 1.82 | 13.8 (12.92–14.38) | 12.62 ± 2.25 | t(42) = -1.79; p = 0.081 | 12.42 ± 1.57 | t(36) = -2.58; p = 0.013 | 12.1 ± 2.02 | t(34) = -2.8; p = 0.092 | ||
Lymphocytes (x10³/mm³) | 1970 ± 485.69 | 1800 (1675–2425) | 2233.33 ± 1241.55 | t(31) = 0.955; p = 0.347 | 2170.83 ± 1019.16 | t(34.2) = 0.856; p = 0.398 | 2398.95 ± 1028.19 | t(34) = 1.815; p = 0.078 | ||
Serum Iron (µg/dL) | 84.1 ± 36.28 | 85.31 (64.51-103.01) | 87.4 ± 38.18 | t(42) = 0.292; p = 0.772 | 95.03 ± 49.7 | t(36) = 0.818; p = 0.418 | 89.77 ± 38 | t(34) = 0.503; p = 0.617 | ||
Ferritin (ng/mL) | 148.34 ± 144.2 | 97.75 (49.42–179.8) | 336 (93.67–828.8) | U = 362; p = 0.004 | 254.6 (159.27–554.3) | U = 359; p = 0.005 | 189.6 (119.4-401.5) | U = 345.000; p = 0.014 | ||
UIBC (µg/dL) | 318.29 ± 39.91 | 320.12 (298.79-339.53) | 287.71 ± 70.77 | t(42) = -1.716; p = 0.247 | 296.85 (276.78-307.78) | U = 153; p = 0.144 | 298.09 ± 36.16 | t(34) = -1.76; p = 0.237 | ||
Transferrin (mg/dL) | 254.63 ± 31.93 | 256.1 (239.03-271.63) | 230.17 ± 56.61 | t(42) = -1.711; p = 0.241 | 237.48 (221.42-246.22) | U = 151; p = 0.140 | 238.47 ± 28.92 | t(34) = -1.71; p = 0.230 | ||
Transferrin Saturation (%) | 26.79 ± 11.76 | 28.6 (19.72–34.5) | 31.45 ± 14.67 | t(42) = 1.146; p = 0.258 | 32.07 ± 15.61 | t(36) = 1.245; p = 0.421 | 30.29 ± 13.2 | t(34) = 0.92; p = 0.363 | ||
Urea (mg/dL) | 33.49 ± 6.99 | 34.16 (28.65-40) | 34.72 (25.59–41.4) | U = 236; p = 0.934 | 31.75 (23.37–38.97) | U = 209; p = 0.472 | 31.25 (21.96–40.44) | U = 202; p = 0.376 | ||
Creatinine (mg/dL) | 1.06 ± 0.23 | 1.12 (0.93–1.26) | 0.92 (0.8–1.16) | U = 189; p = 0.238 | 0.84 (0.74–1.14) | U = 150; p = 0.126 | 0.87 (0.78–1.09) | U = 165.5; p = 0.234 | ||
Aspartate Aminotransferase (AST) (U/L) | 25.21 ± 7.12 | 24.01 (18.96–28.54) | 147.88 (81.54–267.3) | U = 431; p = 0.000 | 29.6 (24.15–38.55) | U = 334; p = 0.027 | 27.85 (25.46–38.81) | U = 165.5; p = 0.081 | ||
Alanine Aminotransferase (ALT) (U/L) | 23.99 ± 10.61 | 24.48 (15.97–29.56) | 146.68 (71.68-253.93) | U = 455; p = 0.000 | 29.73 (16.52–39.14) | U = 310; p = 0.101 | 34.09 (23.5-42.54) | U = 345; p = 0.014 | ||
Folic Acid (ng/mL) | 18.89 ± 4.74 | 19.34 (16.79–22.66) | 13.85 (9.91–18.99) | U = 142; p = 0.021 | 20.67 (10.81-24) | U = 240; p = 1.000 | 21.52 (13.8–24) | U = 253; p = 0.766 | ||
Vitamin B12 (pg/mL) | 497.8 ± 143.75 | 460.5 (417-566.75) | 478 (410–688) | U = 26; p = 0.517 | 561.5 (382.5–1047) | U = 285.5; p = 0.028 | 359.5 (313-544.5) | U = 179; p = 0.091 | ||
Vitamin A (µg/dL) | 0.58 ± 0.16 | 0.55 (0.44–0.71) | 0.30 (0.19–0.35) | U = 28; p = 0.000 | 0.33 ± 0.11 | t(32.9) = -5.853; p = 0.000 | 0.3 (0.26–0.38) | U = 49.5; p = 0.000 | ||
Vitamin E (mg/dL) | 9.05 ± 2.73 | 9.3 (6.53–10.58) | 7.46 (6.11–8.6) | U = 194; p = 0.283 | 7.34 ± 2.41 | t(36) = -2.198; p = 0.034 | 4.98 (4.88–6.13) | U = 74; p = 0.000 | ||
Copper (µg/dL) | 106.83 ± 23.38 | 109 (91.95–113.8) | 139 (105.35-152.45) | U = 351; p = 0.009 | 103.94 ± 31.45 | t(36) = -0.34; p = 0.736 | 95.7 ± 27.61 | t(34) = -1.426; p = 0.161 | ||
Zinc (µg/dL) | 81.92 ± 10.16 | 83.5 (74.57–87.12) | 87.5 (58.25-96) | U = 248; p = 0.860 | 77 (73.62–86.62) | U = 223.5; p = 0.706 | 79.92 ± 16.36 | t(34) = -0.474; p = 0.638 | ||
UIBC: Unsaturated Iron-Binding Capacity; g/dL: grams per deciliter; mg/dL: milligrams per deciliter; mg/L: milligrams per liter; µg/dL: micrograms per deciliter; ng/mL: nanograms per milliliter; pg/mL: picograms per milliliter; x10³/mm³: thousand cells per cubic millimeter; U/L: units per liter; %: percentage. Data are expressed as mean ± SD or median (P25–P75%), according to the statistical distribution observed in each comparison. P-values refer to the statistical models applied for comparisons between time points: p¹ = pre-operative vs. 3 months; p² = 3 months vs. 6 months; p³ = pre-operative vs. 6 months. Mixed-effect linear models (t-test, with degrees of freedom estimated by Satterthwaite's method) were used for variables with an approximately normal distribution, and generalized linear models (z-statistic) for variables with a non-normal distribution, according to the suitability of each variable. The Holm-Bonferroni method was used for the correction of multiple comparisons. P-values ≤ 0.05 were considered statistically significant.. | ||||||||||
Table 6. Linear regression analysis between nutrient intake and serum biomarkers adjusted for CRP (mg/L) and Ferritin (ng/mL) in the PD Group. | ||||||||
|---|---|---|---|---|---|---|---|---|
Preoperative (n = 24) | 3 months after discharge (n = 18) | 5 months after discharge 6 (n = 16) | ||||||
Intake variable | Corresponding biomarker | Adjustment | p | Significance | p | Significance | p | Significance |
Proteins (g) | Albumin (g/dL) | CRP | t(67) = 0.925, p = 0.358 | Not significant | t(67) = 0.925, p = 0.358 | Not significant | t(67) = 0.925, p = 0.358 | Not significant |
Total proteins (g/dL) | CRP | t(67) = 0.795, p = 0.43 | Not significant | t(67) = 0.795, p = 0.43 | Not significant | t(67) = 0.795, p = 0.43 | Not significant | |
Iron (mg) | Serum iron (µg/dL) | CRP | t(67) = 1.53, p = 0.131 | Not significant | t(67) = 1.53, p = 0.131 | Not significant | t(67) = 1.53, p = 0.131 | Not significant |
Ferritin (ng/mL) | CRP | t(67) = 1.966, p = 0.053 | Trend, marginal | t(67) = 1.966, p = 0.053 | Trend, marginal | t(67) = 1.966, p = 0.053 | Trend, marginal | |
Hemoglobin (g/dL) | CRP | t(67) = 2.226, p = 0.029 | Significant | t(67) = 2.226, p = 0.029 | Significant | t(67) = 2.226, p = 0.029 | Significant | |
Zinc (mg) | Serum zinc (µg/dL) | CRP | t(67) = 0.689, p = 0.493 | Not significant | t(67) = 0.689, p = 0.493 | Not significant | t(67) = 0.689, p = 0.493 | Not significant |
Vitamin A (mg) | Vitamin A (µg/dL) | CRP | t(67) = -0.32, p = 0.75 | Not significant | t(67) = -0.32, p = 0.75 | Not significant | t(67) = -0.32, p = 0.75 | Not significant |
Vitamin E (mg) | Vitamin E (mg/dL) | CRP | t(67) = 1.588, p = 0.117 | Not significant | t(67) = 1.588, p = 0.117 | Not significant | t(67) = 1.588, p = 0.117 | Not significant |
Proteins (g) | Albumin (g/dL) | Ferritin | t(67) = 0.338, p = 0.736 | Not significant | t(67) = 0.338, p = 0.736 | Not significant | t(67) = 0.338, p = 0.736 | Not significant |
Total proteins (g/dL) | Ferritin | t(67) = 1.153, p = 0.253 | Not significant | t(67) = 1.153, p = 0.253 | Not significant | t(67) = 1.153, p = 0.253 | Not significant | |
Iron (mg) | Serum iron (µg/dL) | Ferritin | t(67) = 1.42, p = 0.16 | Not significant | t(67) = 1.42, p = 0.16 | Not significant | t(67) = 1.42, p = 0.16 | Not significant |
Ferritin (ng/mL) | Ferritin | t(68) = 2.116, p = 0.038 | Significant | t(68) = 2.116, p = 0.038 | Significant | t(68) = 2.116, p = 0.038 | Significant | |
Hemoglobin (g/dL) | Ferritin | t(67) = 2.027, p = 0.047 | Significant | t(67) = 2.027, p = 0.047 | Significant | t(67) = 2.027, p = 0.047 | Significant | |
Zinc (mg) | Serum zinc (µg/dL) | Ferritin | t(67) = 0.087, p = 0.931 | Not significant | t(67) = 0.087, p = 0.931 | Not significant | t(67) = 0.087, p = 0.931 | Not significant |
Vitamin A (mg) | Vitamin A (µg/dL) | Ferritin | t(67) = -0.245, p = 0.807 | Not significant | t(67) = -0.245, p = 0.807 | Not significant | t(67) = -0.245, p = 0.807 | Not significant |
Vitamin E (mg) | Vitamin E (mg/dL) | Ferritin | t(67) = 1.593, p = 0.116 | Not significant | t(67) = 1.593, p = 0.116 | Not significant | t(67) = 1.593, p = 0.116 | Not significant |
g = gram; mg = milligram; µg = microgram; ng = nanogram; mL = milliliter; dL = deciliter. Linear regression models were adjusted for C-reactive protein (CRP), ferritin, age, and sex. p values refer to the regression coefficient of the corresponding intake variable. * Statistically significant values (p ≤ 0.05). Results were consistent across the three time points evaluated (preoperative, and 3 and 6 months after discharge). | ||||||||
Table 6. Linear regression analysis between nutrient intake and serum biomarkers adjusted for CRP (mg/L) and Ferritin (ng/mL) in the PD Group. | ||||||||
|---|---|---|---|---|---|---|---|---|
Preoperative (n = 24) | 3 months after discharge (n = 18) | 7 months after discharge 8 (n = 16) | ||||||
Intake variable | Corresponding biomarker | Adjustment | p | Significance | p | Significance | p | Significance |
Proteins (g) | Albumin (g/dL) | CRP | t(67) = 0.925, p = 0.358 | Not significant | t(67) = 0.925, p = 0.358 | Not significant | t(67) = 0.925, p = 0.358 | Not significant |
Total proteins (g/dL) | CRP | t(67) = 0.795, p = 0.43 | Not significant | t(67) = 0.795, p = 0.43 | Not significant | t(67) = 0.795, p = 0.43 | Not significant | |
Iron (mg) | Serum iron (µg/dL) | CRP | t(67) = 1.53, p = 0.131 | Not significant | t(67) = 1.53, p = 0.131 | Not significant | t(67) = 1.53, p = 0.131 | Not significant |
Ferritin (ng/mL) | CRP | t(67) = 1.966, p = 0.053 | Trend, marginal | t(67) = 1.966, p = 0.053 | Trend, marginal | t(67) = 1.966, p = 0.053 | Trend, marginal | |
Hemoglobin (g/dL) | CRP | t(67) = 2.226, p = 0.029 | Significant | t(67) = 2.226, p = 0.029 | Significant | t(67) = 2.226, p = 0.029 | Significant | |
Zinc (mg) | Serum zinc (µg/dL) | CRP | t(67) = 0.689, p = 0.493 | Not significant | t(67) = 0.689, p = 0.493 | Not significant | t(67) = 0.689, p = 0.493 | Not significant |
Vitamin A (mg) | Vitamin A (µg/dL) | CRP | t(67) = -0.32, p = 0.75 | Not significant | t(67) = -0.32, p = 0.75 | Not significant | t(67) = -0.32, p = 0.75 | Not significant |
Vitamin E (mg) | Vitamin E (mg/dL) | CRP | t(67) = 1.588, p = 0.117 | Not significant | t(67) = 1.588, p = 0.117 | Not significant | t(67) = 1.588, p = 0.117 | Not significant |
Proteins (g) | Albumin (g/dL) | Ferritin | t(67) = 0.338, p = 0.736 | Not significant | t(67) = 0.338, p = 0.736 | Not significant | t(67) = 0.338, p = 0.736 | Not significant |
Total proteins (g/dL) | Ferritin | t(67) = 1.153, p = 0.253 | Not significant | t(67) = 1.153, p = 0.253 | Not significant | t(67) = 1.153, p = 0.253 | Not significant | |
Iron (mg) | Serum iron (µg/dL) | Ferritin | t(67) = 1.42, p = 0.16 | Not significant | t(67) = 1.42, p = 0.16 | Not significant | t(67) = 1.42, p = 0.16 | Not significant |
Ferritin (ng/mL) | Ferritin | t(68) = 2.116, p = 0.038 | Significant | t(68) = 2.116, p = 0.038 | Significant | t(68) = 2.116, p = 0.038 | Significant | |
Hemoglobin (g/dL) | Ferritin | t(67) = 2.027, p = 0.047 | Significant | t(67) = 2.027, p = 0.047 | Significant | t(67) = 2.027, p = 0.047 | Significant | |
Zinc (mg) | Serum zinc (µg/dL) | Ferritin | t(67) = 0.087, p = 0.931 | Not significant | t(67) = 0.087, p = 0.931 | Not significant | t(67) = 0.087, p = 0.931 | Not significant |
Vitamin A (mg) | Vitamin A (µg/dL) | Ferritin | t(67) = -0.245, p = 0.807 | Not significant | t(67) = -0.245, p = 0.807 | Not significant | t(67) = -0.245, p = 0.807 | Not significant |
Vitamin E (mg) | Vitamin E (mg/dL) | Ferritin | t(67) = 1.593, p = 0.116 | Not significant | t(67) = 1.593, p = 0.116 | Not significant | t(67) = 1.593, p = 0.116 | Not significant |
g = gram; mg = milligram; µg = microgram; ng = nanogram; mL = milliliter; dL = deciliter. Linear regression models were adjusted for C-reactive protein (CRP), ferritin, age, and sex. p values refer to the regression coefficient of the corresponding intake variable. * Statistically significant values (p ≤ 0.05). Results were consistent across the three time points evaluated (preoperative, and 3 and 6 months after discharge). | ||||||||
Table 7. Linear regression analysis between nutrient intake and serum biomarkers adjusted for CRP (mg/L) and Ferritin (ng/mL) in the Controle Group. | ||||
|---|---|---|---|---|
Control Group (n = 20) | ||||
Intake variable | Corresponding biomarker | Adjustment | p | Significance |
Proteins (g) | Albumin (g/dL) | CRP | t(15) = -0.63, p = 0.538 | Not significant |
Total proteins (g/dL) | CRP | t(15) = 1.713, p = 0.107 | Not significant | |
Iron (mg) | Serum iron (µg/dL) | CRP | t(15) = -0.738, p = 0.472 | Not significant |
Ferritin (ng/mL) | CRP | t(15) = -0.728, p = 0.478 | Not significant | |
Hemoglobin (g/dL) | CRP | t(15) = 0.813, p = 0.429 | Not significant | |
Zinc (mg) | Serum zinc (µg/dL) | CRP | t(15) = -0.666, p = 0.515 | Not significant |
Vitamin A (mg) | Vitamin A (µg/dL) | CRP | t(15) = -0.55, p = 0.59 | Not significant |
Vitamin E (mg) | Vitamin E (mg/dL) | CRP | t(15) = -1.766, p = 0.098 | Trend, marginal |
Proteins (g) | Albumin (g/dL) | Ferritin | t(15) = -0.85, p = 0.409 | Não significativa |
Total proteins (g/dL) | Ferritin | t(15) = 1.869, p = 0.081 | Trend, marginal | |
Iron (mg) | Serum iron (µg/dL) | Ferritin | t(15) = -0.587, p = 0.566 | Not significant |
Ferritin (ng/mL) | Ferritin | t(16) = -0.76, p = 0.458 | Not significant | |
Hemoglobin (g/dL) | Ferritin | t(15) = 1.15, p = 0.268 | Not significant | |
Zinc (mg) | Serum zinc (µg/dL) | Ferritin | t(15) = -0.778, p = 0.448 | Not significant |
Vitamin A (mg) | Vitamin A (µg/dL) | Ferritin | t(15) = -0.448, p = 0.661 | Not significant |
Vitamin E (mg) | Vitamin E (mg/dL) | Ferritin | t(15) = -2.342, p = 0.033 | Significant |
g = gram; mg = milligram; µg = microgram; ng = nanogram; mL = milliliter; dL = deciliter. Linear regression models were adjusted for C-reactive protein (CRP), ferritin, age, and sex. p values refer to the regression coefficient of the corresponding intake variable. * Statistically significant values (p ≤ 0.05). Results were consistent across the three time points evaluated (preoperative, and 3 and 6 months after discharge). | ||||
In the pre-operative period, patients exhibited higher levels of C-reactive protein (p < 0.001), AST (p < 0.001), ALT (p < 0.001), ferritin (p = 0.004), and copper (p = 0.009), along with lower concentrations of folic acid (p = 0.021) and vitamin A (p < 0.001). There were no significant differences for the other laboratory markers.
At three months post-discharge, a reduction in albumin (p = 0.015) and hemoglobin (p = 0.013) levels was observed, with a maintained elevation of C-reactive protein (p = 0.010) and an increase in AST (p = 0.027), with no significant difference for ALT (p = 0.101). There was also a reduction in vitamins A (p < 0.001) and E (p = 0.034), along with an increase in vitamin B12 (p = 0.028). During this period, the other parameters showed no significant changes.
At six months post-discharge, ferritin levels remained higher in the surgical group (p = 0.014), while vitamins A and E remained reduced (both p < 0.001). ALT also showed a significant elevation (p = 0.014). The other variables, including hemoglobin (p = 0.092), did not differ between the groups.
4.4 Association between Food Intake and Clinical-Nutritional Biomarkers
Table 6 presents the results of the multiple linear regression between nutrient intake and serum biomarkers in the PD group, with adjustments for C-reactive protein, ferritin, age, and sex. No statistically significant associations were observed between protein intake and the biomarkers albumin (p = 0.358) and total proteins (p = 0.430), nor between the micronutrients zinc, vitamin A, and vitamin E and their respective biomarkers (p > 0.05). Iron intake, however, showed a significant association with hemoglobin (p = 0.029), regardless of the time point analyzed. When adjusted for ferritin, iron intake was also significantly associated with ferritin (p = 0.038) and hemoglobin (p = 0.047), suggesting a possible positive influence of intake on these markers. There was also a marginal trend for an association between iron intake and ferritin (p = 0.053) in the model adjusted for CRP.
Table 7 shows the results of the same analysis applied to the control group. No significant association was found between protein intake and the biomarkers albumin (p = 0.538) or total proteins (p = 0.107), nor between iron intake and the markers serum iron, ferritin, and hemoglobin (p > 0.05 for all). No significant associations were observed between the intake of zinc, vitamin A, and vitamin E and their respective biomarkers, with the exception of a trend between vitamin E intake and its serum levels adjusted for CRP (p = 0.098). After adjustment for ferritin, vitamin E intake showed a statistically significant association with its plasma concentration (p = 0.033), while the other variables remained without statistical significance.
5. Discussion
This prospective, observational, longitudinal, and controlled cohort investigated patients who underwent pancreatoduodenectomy (PD) for periampullary malignant neoplasms. The study evaluated food intake and clinical-nutritional biomarkers in the pre-operative period, and at 3 and 6 months post-discharge.
A
The objectives included describing and comparing the evolution of food intake and biomarkers relative to a control group, and analyzing correlations to support individualized nutritional strategies.5.1 Sample Characteristics
The analysis of clinical and demographic characteristics demonstrated a predominance of the male sex in the group that underwent pancreatoduodenectomy, which is in line with epidemiological data on periampullary neoplasms [35]. The age homogeneity, with an average above 60 years, reinforces the validity of the sample, as aging is related to an increased risk of neoplasms and changes in energy metabolism [36]. The anatomical distribution of tumors, with a predominance of pancreatic head adenocarcinomas (45.8%) and ampulla of Vater (33.3%), aligns with the literature, as these locations are frequently associated with early obstructive symptoms and diagnosis at resectable stages [37]. The inclusion of less prevalent neoplasms broadened the clinical representativeness of the sample. It was also observed that most patients had disease in intermediate stages (pT2 or pT3), and 21% (5/24) had lymph node involvement (N1/N2). Lymph node involvement is associated with immunometabolic activation and systemic inflammation, resulting in a more compromised metabolic phenotype [38, 39]. These findings, in clinical and nutritional practice, reinforce the importance of integrated early screening and intervention strategies, taking into account tumor staging and the metabolic profile [40].
5.2 Food Intake
The longitudinal evaluation of food intake in patients who underwent pancreatoduodenectomy (PD) evidenced persistent inadequacy of energy and protein intake over six months of follow-up. A marked reduction in caloric intake was observed after hospital discharge, with an average below 2100 kcal, in contrast to values exceeding 2700 kcal in the control group (p < 0.0001). This deficit, which was statistically significant and maintained over time, places these patients below the recommended 25–30 kcal/kg/day advocated by ESPEN [41]. Protein consumption remained below the recommended minimum of 1.2 g/kg/day, a situation that can be aggravated by exocrine pancreatic insufficiency, which compromises the digestion and absorption of proteins [42, 43]. The low intake of these macronutrients compromises the preservation of lean body mass, interferes with the adaptive immunometabolic response, and increases the risk of infectious complications, delayed wound healing, and prolonged convalescence [44, 45]. Factors such as anorexia, gastroparesis, dysbiosis, and malabsorption contribute to this scenario, making early nutritional support essential [46].
Deficiencies in micronutrient intake were also observed, especially in iron, calcium, selenium, and vitamin E, with no significant recovery over time (p < 0.001), consistent with reports of deficiency after duodenal resection and lipid malabsorption [47, 48]. Potassium and sodium intake was lower, but this may reflect reduced caloric intake more than a specific absorptive deficiency [47]. Notably, zinc intake, although elevated pre-operatively, remained above the RDA after six months, and vitamin C intake was significantly higher than in the control group at all time points [48]. On the other hand, vitamin A also remained above the RDA even after surgery, in contrast to classic deficiency reports [47], likely reflecting specific dietary patterns of the studied population.
A
Despite the reduction, fiber intake remained adequate or above the recommendation (25–38 g/day) in the post-operative periods, and was higher than in the control group, which may reflect the maintenance of previous habits or partial adherence to nutritional guidelines. High-fiber diets are beneficial for the microbiota, motility, and gastrointestinal symptoms after PD, especially when associated with the correct use of pancreatic enzymes [49]. Therefore, it was documented that patients who underwent PD maintain food intake below the minimum required for metabolic and functional recovery, emphasizing the limitation of conventional hospital discharge approaches. This reinforces the need for structured nutritional rehabilitation programs, with oral and/or enteral support, enzyme replacement, and multi-professional follow-up [50, 51]. The absence of targeted interventions contributes to a chronic catabolic state, increased hospitalizations, worse quality of life, and reduced survival. Thus, clinical nutrition should be central to post-pancreatoduodenectomy oncological care, figuring as a priority therapeutic intervention.5.3 Clinical-Nutritional Biomarkers
The presence of a low-grade inflammatory state in the PD group is a finding that deserves attention. C-reactive protein (CRP), although showing a significant reduction after surgery (p < 0.0001), remained elevated even at six months (p = 0.001). This subclinical inflammation affects metabolic, protein, and immunological dysfunctions, in addition to anabolic resistance [52, 53]. In parallel, serum albumin showed a downward trend in the PD group, with a significant difference at three months (p = 0.015), suggesting a compromise of functional protein reserve, which is associated with a worse prognosis [53]. The slight reduction in total proteins reinforces the interaction between persistent inflammation, reduced hepatic synthesis, and inadequate intake [54]. The combination of CRP with albumin can optimize risk stratification [52].
In addition to the inflammatory profile, relevant changes in absorption markers and micronutrients were also identified during follow-up. The progression of steatorrhea (a reduction in negative tests from 72.2% to 56.25% at six months) was accompanied by deficiencies of vitamins A and E (p < 0.005), highlighting exocrine pancreatic insufficiency [55, 47]. The reduction in copper (p < 0.0001) and zinc values close to the lower limit suggest absorptive and hepatic dysfunction, with an impact on immunity [56]. In the hematological context, a significant reduction in hemoglobin was verified at three months (p = 0.013), a finding consistent with multifactorial anemia [39], while lymphocytic stability suggests the preservation of immunological competence [57]. The significant reduction in ferritin (p < 0.005), although maintained at high levels, can mask functional deficiencies, which reinforces the importance of joint analysis with transferrin and serum iron for better diagnostic accuracy [58].
Regarding metabolic and organic functions, renal function remained preserved [59], and blood glucose showed stability throughout the follow-up, although continuous monitoring is recommended due to the potential impact on tumor evolution [60, 61]. There was a significant improvement in AST and ALT (p < 0.0001), suggesting progressive functional recovery [62, 63]. The increase in vitamin B12 (p < 0.05) in the PD group is possibly related to hepatic dysfunction, and screening for holotranscobalamin and homocysteine is indicated, as high levels have been associated with a worse oncological prognosis [64].
5.4 Association between Food Intake and Clinical-Nutritional Biomarkers
Despite the widespread use of serum biomarkers in the nutritional assessment of surgical patients, our findings confirm the limitation of these indicators as a direct reflection of food intake, in line with previous studies that point to a low correlation between diet and biomarkers in populations undergoing major abdominal surgeries [65, 54]. The lack of a significant correlation between protein intake and serum levels of albumin and total proteins highlights the multifactorial nature of these biomarkers, which are often influenced by acute inflammatory response, hydroelectrolytic status, and liver function [66, 65]. This interpretation is endorsed by ESPEN guidelines, which recommend caution when using albumin as a marker of nutritional status, emphasizing its primary role as an acute-phase reactant [66].
The fact that iron intake was a significant predictor of hemoglobin and ferritin levels in the surgical group suggests that, in contexts of high metabolic demand, adequate micronutrient consumption can be decisive for the prevention of specific deficiencies, corroborating the results of previous studies in at-risk populations [54]. In the control group, dietary intake was a significant predictor of serum vitamin E levels, which is consistent with literature that highlights greater sensitivity of nutritional biomarkers in clinical situations with less inflammatory stress [65]. These findings reinforce the need for a multiparametric approach to nutritional assessment, where the interpretation of biomarkers is always contextualized to the clinical picture, the presence of inflammation, and possible confounding factors [66, 65, 54]. Therefore, the isolated analysis of the relationship between diet and biomarkers may be insufficient to guide clinical decisions, and the integration of multiple parameters and updated recommendations is recommended.
5.5 Study Limitations and Strengths
5.5.1 Limitations
This study, although relevant, presents limitations that should be considered in the interpretation of its findings. The sample size, although compatible with the target population, may have reduced the ability to detect associations of smaller magnitude. Follow-up losses, mainly due to death, generated missing data at the 3 and 6-month time points; although multiple imputation by chained equations (MICE) was applied, the risk of residual bias cannot be completely ruled out. Dietary assessment, based on a self-reported Food Frequency Questionnaire (FFQ), is subject to recall bias, which was minimized by its application by a trained team. As this was a study conducted in a single center, the generalization of the results should be done with caution, considering differences in care and population between services.
5.5.2 Strengths
The prospective, longitudinal, and controlled design is uncommon in the literature on highly complex oncological surgeries and strengthens the robustness of the findings. The integration of food intake data with clinical-laboratory biomarkers, analyzed by multiple linear regression models adjusted for inflammatory markers (CRP and ferritin), age, and sex, made it possible to control for potential confounding factors. Multiple imputation by chained equations (MICE), combined according to Rubin's rules, reduced information loss and increased the precision of the estimates. The collection of dietary data through an adapted 114-item FFQ, applied by a trained and stable multi-professional team, ensured methodological uniformity. The study's conduct in a specialized center, with standardized surgical and care protocols, reinforced the internal consistency and reproducibility of the results in similar contexts. Furthermore, the present study fully complied with the recommendations of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement, ensuring transparency, standardization, and completeness in the presentation of the findings.
6. Conclusion
The present study demonstrated that patients with periampullary malignant neoplasms who underwent pancreatoduodenectomy show a significant reduction in energy and protein intake, associated with persistent systemic inflammation and alterations in nutritional biomarkers during the six months following surgery. Deficiencies of fat-soluble vitamins (A and E) and minerals (iron, zinc, and copper) were observed, as well as a decrease in albumin and total protein levels, reflecting the complex interaction between inadequate food consumption, inflammatory response, and exocrine pancreatic insufficiency.
Although most biomarkers did not show a strong correlation with dietary intake, the longitudinal analysis indicated that iron intake was a predictor of hemoglobin and ferritin levels in the surgical group, while vitamin E intake influenced its serum levels in the control group. These findings highlight the relevant role of specific micronutrients in post-operative recovery, reinforcing their importance in individualized nutritional follow-up.
Thus, it is concluded that pancreatoduodenectomy has a substantial impact on the nutritional status of these patients, highlighting the need for rigorous multiparametric monitoring and early and individualized nutritional interventions, starting in the perioperative period. The adoption of standardized and evidence-based nutritional protocols can lead to better clinical outcomes and a reduction in complications. However, new prospective, longitudinal, controlled, and multicenter studies are essential to deepen the understanding of nutritional changes in this context, promoting continuous advancements in the care of patients who undergo pancreatic surgery.
7. Declaration of artificial intelligence and digital tools use
The authors declare that they used digital tools to support the preparation of this manuscript. STORM (Stanford OVAL Lab, Stanford University, 2024) was used in the brainstorming stage for the conception of the research project. SciSpace (Milpitas, California, USA) assisted in the search for scientific articles related to the study's theme. NotebookLM (Google, Mountain View, California, USA) was used to organize summaries. EndNote (Clarivate, Philadelphia, PA, USA) was used to organize the bibliographic references. Grammarly (San Francisco, California, USA) was applied for English grammatical correction. All analytical, conceptual, and interpretive decisions were conducted by the authors, who assume full responsibility for the scientific content of the article.
Conflict of Interest
The authors declare that there are no conflicts of interest related to this work.
A
Funding
This study was supported by a Master’s scholarship from the Coordination for the Improvement of Higher Education Personnel (CAPES – Finance Code 001). This financial support was strictly intended for student subsistence and did not cover research-related expenses. The funding source had no role in the study design, data collection, analysis, or interpretation, nor in the writing of the manuscript.
A
Author Contribution
1.Marco Aurélio Ribeiro (MAR) was responsible for the conception of the study, data collection, statistical analysis, and drafting of the manuscript.2.José Sebastião dos Santos (JSS) supervised the technical aspects related to gastrointestinal surgery and critically revised the manuscript.3.Gleici da Silva Perdoná (GSP) performed statistical analysis, contributed to data interpretation, and revised the manuscript.4.Anderson Marliere Navarro (AMN), as the academic supervisor of the study, provided methodological support, overall guidance, data interpretation, and critical revision of the manuscript.
1.
José Sebastião dos Santos (JSS) supervised the technical aspects related to gastrointestinal surgery and critically revised the manuscript.
2.
Gleici da Silva Perdoná (GSP) performed statistical analysis, contributed to data interpretation, and revised the manuscript.
3.
A
Anderson Marliere Navarro (AMN), as the academic supervisor of the study, provided methodological support, overall guidance, data interpretation, and critical revision of the manuscript.A
Acknowledgement
A
The authors would like to thank the staff of the Hospital das Clínicas of the Ribeirão Preto Medical School, University of São Paulo, for their support in data collection and patient care. We are also grateful to the patients and their families for their participation and trust.A
Data Availability
The datasets generated and/or analyzed during the current study are not publicly available due to patient confidentiality and ethical restrictions, but may be made available by the corresponding author upon reasonable request and with approval from the Research Ethics Committee.
Reference
1.
Moore A, Donahue T. Pancreatic cancer. JAMA. 2019;322(14):1426. doi:10.1001/jama.2019.14699.
2.
Tosti L, Hang Y, Debnath O, et al. Single-nucleus and in situ RNA-sequencing reveal cell topographies in the human pancreas. Gastroenterology. 2021;160(4):1330-44.e11. doi:10.1053/j.gastro.2020.11.010.
3.
Li T, Lin C, Wang W. Global, regional, and national burden of pancreatic cancer from 1990 to 2021, its attributable risk factors, and projections to 2050: a systematic analysis. BMC Cancer. 2025;25:189. doi:10.1186/s12885-025-13597-z.
A
4.Halbrook CJ, Lyssiotis CA, Pasca di Magliano M, et al. Pancreatic cancer: advances and challenges. Cell. 2023;186(8):1729–54. doi:10.1016/j.cell.2023.02.014.
5.
Gaspar AF, Kemp R, Sankarankutty AK, Júnior JRL, Filho JAF, Martone D. Clinical outcomes and costs of the use of fibrin sealant in pancreaticojejunal anastomosis after pancreaticoduodenectomy: a retrospective analysis study. Ann Med Surg (Lond). 2021;67:102531. doi:10.1016/j.amsu.2021.102531.
6.
Utsumi M, Aoki H, Nagahisa S, et al. Preoperative nutritional assessment using the Controlling Nutritional Status score. In Vivo. 2020;34(4):1931–9. doi:10.21873/invivo.11990.
A
7.Rykina-Tameeva N, Samra JS, Sahni S, et al. Drain fluid biomarkers in pancreatic fistula: a narrative review. World J Gastrointest Surg. 2022;14(10):1089–106. doi:10.4240/wjgs.v14.i10.1089.
8.
Kroon VJ, Daamen LA, Tseng DSJ, et al. Pancreatic exocrine insufficiency after pancreatoduodenectomy. Pancreatology. 2022;22(7):1020–7. doi:10.1016/j.pan.2022.08.002.
9.
Dominguez-Muñoz JE, Vujasinovic M, de la Iglesia D, et al. Guidelines on pancreatic exocrine insufficiency. United European Gastroenterol J. 2025;13(2):125–72. doi:10.1002/ueg2.12674.
10.
Chang VT, Sandifer C, Zhong F. GI symptoms in pancreatic cancer. Clin Colorectal Cancer. 2023;22(1):24–33. doi:10.1016/j.clcc.2022.12.002.
11.
Mękal D, Sobocki J, Badowska-Kozakiewicz A, et al. Nutritional treatment in pancreatic cancer. Cancers (Basel). 2023;15(15):3816. doi:10.3390/cancers15153816.
12.
Kadakia KC, Hamilton-Reeves JM, Baracos VE. Therapeutic targets in cancer cachexia. Am Soc Clin Oncol Educ Book. 2023;43:e389942. doi:10.1200/EDBK_389942.
A
13.Kim E, Lee D-H, Jang J-Y. Preoperative malnutrition and outcomes in periampullary neoplasms. Gut Liver. 2019;13(6):690–7. doi:10.5009/gnl18469.
14.
Muscaritoli M, Arends J, Bachmann P, et al. ESPEN practical guideline: clinical nutrition in cancer. Clin Nutr. 2021;40(5):2898–913. doi:10.1016/j.clnu.2021.02.005.
15.
Poulia KA, Sarantis P, Antoniadou D, et al. Pancreatic cancer and cachexia mechanisms. Nutrients. 2020;12(6):1543. doi:10.3390/nu12061543.
16.
Abu-Khudir R, Hafsa N, Badr BE. Biomarkers for pancreatic cancer prognosis. Diagnostics (Basel). 2023;13(19):3091. doi:10.3390/diagnostics13193091.
A
17.Russell T. The nutritional management of pancreatic resection patients is highly variable [Internet]. Springer Nature Communities; 2022 [cited 2025 Jul 3]. Available from: https://communities.springernature.com/posts/the-nutritional-management-of-pancreatic-resection-patients-is-highly-variable-is-it-time-for-consensus-guidelines.
18.
Martin D, Joliat G-R, Halkic N, et al. Perioperative nutritional management of patients undergoing pancreatoduodenectomy. HPB (Oxford). 2019;22(1):75–82. doi:10.1016/j.hpb.2019.05.009.
A
19.Arends J, Bachmann P, Baracos V, et al. ESPEN guidelines on nutrition in cancer. Clin Nutr. 2017;36(1):11–48. doi:10.1016/j.clnu.2016.07.015.
20.
Ichida H, Imamura H, Takahashi A, et al. Pancreatic morphometry and nutrition post-pancreatectomy. Surgery. 2024;176(9):1189–97. doi:10.1016/j.surg.2024.05.046.
21.
Jotheeswaran R, Singh H, Kaur J, et al. Role of inflammatory and nutritional markers in predicting complications after pancreaticoduodenectomy. Surgery. 2022;172(6):1502–9. doi:10.1016/j.surg.2022.07.002.
22.
Shiihara M, Higuchi R, Izumo W, et al. Impact of nutritional status on complications of pancreaticoduodenectomy. Langenbecks Arch Surg. 2021;406(5):1491–8. doi:10.1007/s00423-021-02151-7.
23.
Halle-Smith JM, Powell-Brett SF, Hall LA, et al. Recent advances in pancreatic ductal adenocarcinoma: strategies to optimise the perioperative nutritional status in pancreatoduodenectomy patients. Cancers (Basel). 2023;15(9):2466. doi:10.3390/cancers15092466.
24.
Maehira H, Iida H, Mori H, et al. Clinical impact of postoperative malnutrition after pancreaticoduodenectomy. HPB (Oxford). 2023;25(9):1102–9. doi:10.1016/j.hpb.2023.04.020.
25.
Sun C-Y, Zhang X-J, Li Z, et al. Preoperative prognostic nutritional index predicts long-term outcomes of patients with ampullary adenocarcinoma. World J Gastrointest Surg. 2024;16(5):1291–300. doi:10.4240/wjgs.v16.i5.1291.
A
26.Molina M del CB, Benseñor IM, Cardoso L de O, et al. Reprodutibilidade e validade relativa do Questionário de Frequência Alimentar do ELSA-Brasil. Cad Saude Publica. 2013;29(2):379–89. doi:10.1590/S0102-311X2013000200024.
A
27.Vannucchi H, De Unamuno M do RDL, Marchini JS. Avaliação do estado nutricional [Internet]. São Paulo: Core; 1996 [cited 2025 Jul 3]. Available from: https://revistas.usp.br/rmrp/article/view/707/719.
28.
White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377–99. doi:10.1002/sim.4067.
29.
Grund S, Lüdtke O, Robitzsch A. On the treatment of missing data in background questionnaires in educational large-scale assessments: an evaluation of different procedures. J Educ Behav Stat. 2021;46(4):430–65. doi:10.3102/1076998620959058.
30.
Rubin DB. Inference and missing data. Biometrika. 1976;63(3):581–95. doi:10.1093/biomet/63.3.581.
31.
Lu S-E, Kim S, Cheng JQ, et al. Fixed and random effect selections in generalized linear mixed models. Stat Methods Med Res. 2024;33(1):3–23. doi:10.1177/09622802231221201.
32.
Casals M, Girabent-Farrés M, Carrasco JL. Methodological quality and reporting of generalized linear mixed models in clinical medicine (2000–2012): a systematic review. PLoS One. 2014;9(11):e112653. doi:10.1371/journal.pone.0112653.
33.
Hadfield JD. MCMC methods for multi-response generalized linear mixed models: the MCMCglmm R package. J Stat Softw. 2010;33(2):1–22. doi:10.18637/jss.v033.i02.
34.
Amin MB, Greene FL, Edge SB, et al. The Eighth Edition AJCC Cancer Staging Manual: continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging. CA Cancer J Clin. 2017;67(2):93–9. doi:10.3322/caac.21388.
35.
von Fritsch L, Duhn J, Abdalla TSA, et al. An R0 resection margin does improve overall survival after PDAC resection—real-world evidence from 6000 cases from the German Cancer Registry Group. Eur J Surg Oncol. 2025;51(7):109693. doi:10.1016/j.ejso.2025.109693.
36.
Cruz-Jentoft AJ, Bahat G, Bauer J, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48(1):16–31. doi:10.1093/ageing/afy169.
37.
Kaluba B, Kuriyama N, Sakamoto T, et al. Long-term survival analysis based on tumor location in patients with pancreatic ductal adenocarcinoma who underwent pancreatectomy following neoadjuvant chemoradiotherapy. Langenbecks Arch Surg. 2025;410:47. doi:10.1007/s00423-025-03609-8.
38.
Baracos VE, Martin L, Korc M, Guttridge DC, Fearon KCH. Cancer-associated cachexia. Nat Rev Dis Primers. 2018;4:17105. doi:10.1038/nrdp.2017.105.
39.
Xu Z, Chen J, Mou Y, et al. Effects of nutritional status on short-term prognosis after minimally invasive pancreaticoduodenectomy. Sci Rep. 2024;14:29549. doi:10.1038/s41598-024-81016-7.
40.
Prado CM, Purcell SA, Alish C, et al. Nutrition and sarcopenia in cancer: importance of a multimodal approach. Curr Opin Clin Nutr Metab Care. 2021;24(5):396–403. doi:10.1097/MCO.0000000000000759.
41.
Weimann A, Braga M, Carli F, Higashiguchi T, Hübner M, Klek S, et al. ESPEN practical guideline: Clinical nutrition in surgery. Clin Nutr. 2021;40(7):4745–61. doi:10.1016/j.clnu.2021.03.031.
42.
Petzel MQB, Hoffman L. Nutrition Implications for Long-Term Survivors of Pancreatic Cancer Surgery. Nutr Clin Pract. 2017;32(5):588–98. doi:10.1177/0884533617722929.
43.
Park S, Lee Y, Kim JY, et al. Management of pancreatic exocrine insufficiency following pancreatoduodenectomy: a multicenter prospective study. Pancreatology. 2025;25:99–107. doi:10.1016/j.pan.2025.02.004.
44.
Martin D, Joliat G-R, Halkic N, et al. Perioperative nutritional management of patients undergoing pancreatoduodenectomy. HPB (Oxford). 2019;22(1):75–82. doi:10.1016/j.hpb.2019.05.009.
45.
Bouloubasi Z, Karayiannis D, Pafili Z, et al. Re-assessing the role of peri-operative nutritional therapy in patients with pancreatic cancer undergoing surgery: a narrative review. Nutr Res Rev. 2024;37(1):121–30. doi:10.1017/S0954422423000100.
46.
Chen D, Li Z, Wang Y, et al. Postoperative complications and nutritional therapy after pancreatic surgery. Front Nutr. 2024;11:1210045. doi:10.3389/fnut.2024.1210045.
47.
Azzini E, Maiani G, Polito A. Vitamin Nutritional Status in Patients with Pancreatic Cancer: A Narrative Review. Int J Mol Sci. 2024;25(9):4773. doi:10.3390/ijms25094773.
48.
Dresen S, Lanz M, Thomé C, et al. Micronutrient deficiencies after pancreatic surgery: the role of inflammation and malabsorption. Clin Nutr. 2023;42(1):64–71. doi:10.1016/j.clnu.2022.08.006.
49.
Nucci G, Salvio G, Altomare A, Emerenziani S, Cicala M. Dietary Fibre Intake and Risk of Pancreatic Cancer: Systematic Review and Meta-Analysis of Observational Studies. Int J Environ Res Public Health. 2021;18(21):11556. doi:10.3390/ijerph182111556.
50.
Mękal D, Sobocki J, Badowska-Kozakiewicz A, et al. Nutritional treatment in pancreatic cancer. Cancers (Basel). 2023;15(15):3816. doi:10.3390/cancers15153816.
51.
Bosch JA, van Dijk DPJ, Slieker JC, et al. Postoperative recovery and nutrition in patients with pancreatic cancer: a prospective study. Support Care Cancer. 2023;31(5):2557–66. doi:10.1007/s00520-023-07882-z.
52.
Shirakawa T, Makiyama A, Shimokawa M, et al. C-reactive protein/albumin ratio is the most significant inflammatory marker in unresectable pancreatic cancer treated with FOLFIRINOX or gemcitabine plus nab-paclitaxel. Sci Rep. 2023;13:8815. doi:10.1038/s41598-023-34962-7.
53.
Tang Q, Li X, Sun C-R. Predictive value of serum albumin levels on cancer survival: a prospective cohort study. Front Oncol. 2024;14:1323192. doi:10.3389/fonc.2024.1323192.
54.
Boyd LNC, Ali M, Leeflang MMG, et al. Diagnostic accuracy and added value of blood-based protein biomarkers for pancreatic cancer: a meta-analysis of aggregate and individual participant data. EClinicalMedicine. 2023;55:101747. doi:10.1016/j.eclinm.2022.101747.
55.
Phillips ME, Hopper AD, Leeds JS, et al. Consensus for the management of pancreatic exocrine insufficiency: UK practical guidelines. BMJ Open Gastroenterol. 2021;8(1):e000643. doi:10.1136/bmjgast-2021-000643.
56.
Zhu B, Huo R, Zhi Q, et al. Increased expression of zinc transporter ZIP4, ZIP11, ZnT1, and ZnT6 predicts poor prognosis in pancreatic cancer. J Trace Elem Med Biol. 2021;65:126734. doi:10.1016/j.jtemb.2021.126734.
57.
Aquilani R, Brugnatelli S, Maestri R, et al. Peripheral blood lymphocyte percentage may predict chemotolerance and survival in patients with advanced pancreatic cancer. Curr Oncol. 2021;28(5):3280–96. doi:10.3390/curroncol28050285.
58.
Ramírez-Carmona W, Díaz-Fabregat B, Yoshigae AY, et al. Are serum ferritin levels a reliable cancer biomarker? A systematic review and meta-analysis. Nutr Cancer. 2022;74(11):1917–26. doi:10.1080/01635581.2021.1982996.
59.
Tsukagoshi M, Watanabe A, Araki K, et al. Usefulness of serum creatinine-to-cystatin C ratio as a screening tool for predicting prognosis in patients with pancreatic cancer. Ann Gastroenterol Surg. 2023;7(6):784–92. doi:10.1002/ags3.12671.
60.
Liu Z, Hayashi H, Matsumura K, et al. Biological and clinical impacts of glucose metabolism in pancreatic ductal adenocarcinoma. Cancers (Basel). 2023;15(2):498. doi:10.3390/cancers15020498.
61.
Vaziri-Gohar A, Hue JJ, Abbas A, et al. Increased glucose availability sensitizes pancreatic cancer to chemotherapy. Nat Commun. 2023;14:3823. doi:10.1038/s41467-023-38921-8.
62.
Chen D, Li Z, Wang Y, Sun C, Zheng S. AST/ALT ratio as a prognostic factor in pancreatic cancer. Pancreas. 2022;51(2):123–9. doi:10.1097/MPA.0000000000001988.
63.
He M, Liu Y, Huang H, et al. Serum aspartate aminotransferase is an adverse prognostic indicator for patients with resectable pancreatic ductal adenocarcinoma. Lab Med. 2023;54(6):608–12. doi:10.1093/labmed/lmad014.
64.
Amado-Garzon SB, Molina-Pimienta L, Vejarano-Pombo A, et al. Elevated vitamin B12, risk of cancer, and mortality: a systematic review. Cancer Invest. 2024;42(7):515–26. doi:10.1080/07357907.2024.2366907.
65.
Zhou S, Yu Z, Shi X, et al. The relationship between phase angle, nutrition status, and complications in patients with pancreatic head cancer. Int J Environ Res Public Health. 2022;19(11):6426. doi:10.3390/ijerph19116426.
66.
Muscaritoli M, Arends J, Bachmann P, et al. ESPEN practical guideline: clinical nutrition in cancer. Clin Nutr. 2021;40(5):2898–913. doi:10.1016/j.clnu.2021.02.005.