Longitudinal assessment of pancreatic exocrine dysfunction in type 1 diabetes

Abstract

BACKGROUND

Previous studies have shown that individuals with type 1 diabetes (T1D) frequently present with reduced fecal elastase levels, suggesting exocrine pancreatic insufficiency. However, the underlying determinants and the longitudinal trajectory of these changes remain poorly understood.

AIM

To evaluate longitudinal changes in fecal elastase among individuals with T1D, identify associated factors, and determine clinical implications.

METHODS

Pancreatic exocrine function was evaluated in a cohort of patients with T1D by measuring fecal elastase concentrations (FECs). After a mean follow-up of 8.5 ± 0.5 years, participants were recontacted, and a second stool sample was obtained. At both time points, detailed medical histories were collected, including information on diabetes progression, metabolic control, complications, gastrointestinal symptoms, and nutritional status. The study was approved by the institutional ethics committee, and written informed consent was obtained from all participants.

RESULTS

A total of 106 individuals with T1D (mean age = 46.2 years; 50% male) were enrolled. At baseline, the median FEC was 239.5 µg/g, with 44 participants (41.5%) demonstrating abnormally low levels (< 200 µg/g). Reduced fecal elastase was significantly associated with male sex, diabetes-related complications, particularly retinopathy, and higher glycated hemoglobin levels. No significant differences in gastrointestinal symptoms, body mass index, nor most serum nutritional markers were observed between individuals with normal vs reduced fecal elastase levels. Sixty-six participants completed follow-up. Their median fecal elastase was 171.5 µg/g, with 59.1% presenting levels below 200 µg/g. Paired analysis showed a non-significant decline in FEC s over time. No clinical nor metabolic variables predicted longitudinal changes in FEC independently.

CONCLUSION

Fecal elastase levels are frequently reduced in individuals with T1D and may show a gradual decline over time. The clinical impact of these changes appears to be limited.

Key Words: Type 1 diabetes; Diabetes-related complications; Pancreatic exocrine function; Fecal elastase-1; Exocrine pancreatic insufficiency

Core Tip: This is the first study to investigate exocrine pancreatic dysfunction in individuals with type 1 diabetes (T1D) using fecal elastase in a longitudinal design. Our findings suggest that pancreatic exocrine function is frequently altered in T1D and may decline over time. However, the underlying pathophysiology remains unclear, and the clinical impact appears limited. Therefore, low fecal elastase levels in T1D should be interpreted with caution to avoid overdiagnosis and unnecessary treatment.

INTRODUCTION

The pancreas is a complex gland with both exocrine and endocrine functions, carried out by distinct cell populations. Rather than operating independently, these cell groups appear to influence one another in ways that are not yet fully understood. Over recent decades, increasing evidence has suggested that the exocrine pancreas may also be involved in diabetes, particularly type 1 diabetes (T1D). Studies have shown that in the T1D pathologic condition the pancreas is smaller[1] and often displays structural changes resembling those observed in chronic pancreatitis[2]. Beyond the alterations in size and morphology, the exocrine function is also impaired. Mohapatra et al[3] termed this constellation of changes as “diabetic exocrine pancreatopathy” in 2016; yet, the underlying pathophysiological mechanisms remain incompletely defined, although several hypotheses have been proposed.

The exocrine function of the pancreas in T1D has been investigated using both direct and indirect tests. Direct tests measure pancreatic secretion but are labor-intensive and rarely applied in routine clinical practice. Based upon these methods, the prevalence of exocrine pancreatic insufficiency (EPI) in T1D has been reported to range from 35% to 77%[4]. Among the indirect methods, the most widely used is that which measures fecal elastase-1. A fecal elastase concentration (FEC) below 200 µg/g is considered diagnostic for EPI, with levels < 100 µg/g typically indicating severe insufficiency. It has been reported that 16% to 57% of patients with T1D have reduced FEC[5,6], though the clinical relevance of these findings remains uncertain.

In a study of 35 patients with T1D, Hahn et al[7] compared FEC with results from the secretin-cerulein test and fecal fat quantification (Van de Kamer method). Correlations were weak, and the authors concluded that low FEC was a poor predictor of EPI in this population. Similar findings were observed in our cohort, where FEC results showed limited agreement with the ¹³C-mixed triglyceride breath test, a practical alternative for measurement of fat malabsorption[8].

The clinical implications of reduced FEC in T1D, the factors contributing to these changes, and their temporal evolution remain poorly understood. Although a progressive decline in pancreatic exocrine function might be expected, this assumption was challenged by a small longitudinal study of 20 patients that found no clear trajectory. Importantly, that study did not account for potential contributing factors (e.g., diabetes-related complications, tobacco or alcohol exposure) nor glycemic control, nor did it examine the clinical impact of the observed changes[9]. Therefore, the aim of our study was to assess longitudinal changes in pancreatic exocrine function in individuals with T1D, identify the associated factors, and evaluate their clinical significance.

MATERIALS AND METHODS

Study design

This was a two-phase cross-sectional study with data collected at two distinct time points. The original cohort included 106 patients with T1D who were referred to the Department of Gastroenterology at the University Hospital of Navarre between 2013 and 2015. All participants were over 18 years of age and met a combination of clinical and laboratory diagnostic criteria, including early age at diagnosis, ketoacidosis at disease onset, early insulin requirement, low C-peptide levels, and the presence of autoantibodies against glutamic acid decarboxylase and tyrosine phosphatase IA-2[10]. Patients with a history of pancreatic disease, active gastrointestinal disorders, or prior abdominal surgery were excluded.

Clinical and laboratory data

A detailed medical history was obtained, including alcohol and tobacco use, as well as gastrointestinal symptoms such as dyspepsia, abdominal pain, or diarrhea. Diabetes-related data were also collected, including disease duration, glycated hemoglobin (HbA1c) levels, and the presence of macrovascular and microvascular complications as indicators of disease control. FEC was measured in all participants using the ScheBo Pancreatic Elastase 1 ELISA kit (Giessen, Germany).

Participants also underwent a comprehensive serum nutritional assessment, which included hemoglobin, mean corpuscular volume, lymphocyte count, prothrombin activity, total serum proteins, albumin, prealbumin, retinol binding protein (RBP), transferrin, ferritin, cholesterol, triglycerides, folic acid, cobalamin, vitamins A, D, and E, phosphate, calcium, iron, zinc, and magnesium. Normal reference ranges for these analytes are provided in Supplementary Table 1.

Follow-up analysis

After a mean follow-up period of 8.5 ± 0.5 years, participants were recontacted and invited to join the second phase of the study. Thirty (28.3%) declined participation, four had died, two had relocated to other cities, and one could not be reached. Two additional patients were no longer eligible due to prior pancreatic surgery for intraductal papillary mucinous neoplasm and neuroendocrine tumor, and one patient was excluded following a right hemicolectomy for colorectal cancer.

The remaining 66 patients (62.3%) participated in the second phase of the study. These individuals were re-evaluated for gastrointestinal symptoms, diabetes-related complications, and exposure to tobacco and alcohol, with medical histories updated accordingly. A second comprehensive assessment of nutritional status was conducted, and a fecal elastase sample was obtained from all participants.

Ethics

The study was approved by the Institutional Review Board (Comité Ético de Experimentación Clínica de Navarra: PI_2013/8) and was conducted in accordance with the ethical guidelines of the 1975 Declaration of Helsinki. Given the two-phase design, separate written informed consent was obtained from all participants prior to each phase.

Statistical analysis

Demographic and clinical characteristics were summarized using frequencies and percentages for categorical variables, and either mean ± SD or median with interquartile range for continuous variables, as appropriate. Data normality was assessed using the Kolmogorov-Smirnov test with Lilliefors correction.

Differences between patients with and without fecal elastase deficiency were evaluated using the χ² test and Fisher’s exact test for categorical variables, and the Mann-Whitney U test for continuous variables. Variables significant in univariate analysis were included in a multivariate logistic regression model, with stepwise backward elimination used for variable selection. Only variables remaining significant in the multivariate model were retained in the final model. To assess potential follow-up bias, patients with and without complete follow-up were compared across all measured variables.

Worsening of FEC among patients with normal baseline levels was examined using a mixed-effects logistic regression model with a patient-level random effect to identify factors associated with deficiency at follow-up. All analyses were performed using R v. 4.2 (R Foundation for Statistical Computing, Vienna, Austria). P values < 0.05 were considered statistically significant.

RESULTS

The study included 106 participants with a mean age of 46.2 ± 11.8 years, 50% of whom were male. The mean diabetes duration was 19.3 years, and 42.5% of participants had diabetes-related complications, predominantly retinopathy (39.6%). The average HbA1c was 7.8%. The median FEC was 239.5 µg/g, with 44 patients (41.5%) showing abnormally low levels (< 200 µg/g). Gastrointestinal symptoms were reported by 37.7% of participants, most commonly dyspepsia (31.1%). Complete demographic and clinical characteristics are presented in Table 1.

Table 1 Characteristics and outcomes.

Clinical characteristics and FEC	Time 1, n = 106
Baseline characteristics
Age (year)	46.2 ± 11.8
Male sex	53 (50)
Smoking	29 (27.3)
Alcohol	19 (17.9)
DM characteristics and control
DM duration in years	19.3 ± 9.7
HbA1c (%)	7.8 ± 1
DM-related complications1	45 (42.5)
Retinopathy	42 (39.6)
Kidney disease	18 (17)
Neuropathy	15 (14.2)
Cardiovascular disease	1 (0.9)
Peripheral artery disease	3 (2.8)
Fecal elastase concentration (µg/g), median (IQR)	239.5 (266)
> 200	62 (58.5)
100-200	21 (19.8)
< 100	23 (21.7)
Gastrointestinal symptoms
No	66 (62.3)
Yes2	40 (37.7)
Diarrhea	6 (5.7)
Abdominal pain	5 (4.7)
Dyspepsia	33 (31.1)
Nutritional status
BMI (kg/m2)	25.2 ± 4.4

¹Some patients presented more than one diabetes-related complication.

²Some patients reported more than one symptom.

Data were showed as mean ± SD or n (%). FEC: Fecal elastase concentration; BMI: Body mass index; DM: Diabetes mellitus; HbA1c: Glycated hemoglobin; IQR: Interquartile range (Q1-Q3).

Participants had a mean body mass index (BMI) of 25.2 kg/m². Nutritional deficiencies were most common for prealbumin (49%), RBP (47%), vitamin A (29%), and vitamin D (32.3%). Less frequent deficiencies included cobalamin (13.4%), ferritin (15.1%), and zinc (14.4%), while most other parameters were within normal ranges or showed low deficiency rates. Nutritional status details are presented in Table 2.

Table 2 Baseline nutritional parameters.

Clinical and blood nutritional parameters	Valid cases	mean ± SD	Below normal limit, n (%)
BMI (kg/m2)	106	25.2 ± 4.4	1 (0.9)
Hemoglobin (g/dL)	106	14.1 ± 1.3	4 (3.8)
Mean corpuscular volume (fL)	106	92.6 ± 5.9	2 (1.9)
Lymphocytes (× 109/L)	106	1842 ± 524	1 (0.9)
Prothrombin activity (%)	105	98.3 ± 4	0
Total proteins (g/dL)	106	6.97 ± 0.42	5 (4.7)
Albumin (g/dL)	106	4.38 ± 0.28	0
Prealbumin (mg/dL)	102	20.9 ± 4.7	50 (49)
Retinol-binding protein (mg/dL)	101	3.2 ± 1.2	48 (47)
Transferrin (mg/dL)	103	233 ± 42	6 (5.8)
Ferritin (µg/L)	103	92 ± 88	16 (15.1)
Cholesterol (mg/dL)	106	181 ± 34	1 (0.9)
Triglycerides (mg/dL)	105	68 ± 33	0
Folic acid (µg/L)	104	11.5 ± 4.7	2 (1.9)
Cobalamin (ng/L)	104	444 ± 259	7 (6.7)
Vitamin A (mg/L)	100	0.38 ± 0.17	12 (12)
Vitamin D (ng/L)	99	25 ± 10	32 (32.3)
Vitamin E (mg/dL)	100	1.34 ± 0.34	0
Phosphate (mg/dL)	106	3.6 ± 0.6	5 (4.7)
Calcium (mg/dL)	106	10.2 ± 7.7	1 (0.9)
Iron (mg/dL)	103	84 ± 34	20 (19.4)
Zinc (µg/dL)	97	76 ± 14	6 (6.2)
Magnesium (mg/dL)	101	1.89 ± 0.19	1 (0.9)

Some tests were not available in a subset of patients. BMI: Body mass index.

Factors related to low FEC

Table 3 compares the demographic, clinical and nutritional characteristics of participants with normal and abnormal FEC at baseline. No significant differences were observed in age, smoking status, nor diabetes duration between the two groups. However, participants with abnormal FEC were more often male (61.4% vs 41.9%, P = 0.049), reported higher alcohol consumption (27.3% vs 11.3%, P = 0.035), had more diabetes-related complications (59.1% vs 30.6%, P = 0.004), particularly retinopathy (56.8% vs 27.4%, P = 0.002), and higher HbA1c levels (8.1% vs 7.6%, P = 0.017) compared with those with normal FEC. In multivariate analysis, only male sex [odds ratio (OR) = 3.1 (1.3-8.1), P = 0.013], retinopathy [OR = 2.6 (1.1-6.4); P = 0.032], and poorer metabolic control as measured by HbA1c [OR = 1.8 (1.2-3.1); P = 0.015] remained independently associated with low FEC.

Table 3 Univariate and multivariate analysis of risk factors and clinical characteristics for low fecal elastase concentration.

Variables	Univariate analysis			Multivariate analysis
	Normal FEC, n = 62	Abnormal FEC, n = 44	P value	OR	95%CI	P value
Age (year)	45.1 ± 11.7	47.3 ± 11.8	0.364
Male sex	26 (41.9)	27 (61.4)	0.049	3.14	1.30-8.06	0.013
Female sex	36 (58.1)	17 (38.6)
Smoking	13 (21)	16 (36.4)	0.080
Alcohol	7 (11.3)	12 (27.3)	0.035
DM duration (year)	18.4 ± 10	20.8 ± 10.8	0.244
DM-related complications	19 (30.6)	26 (59.1)	0.004
Retinopathy	17 (27.4)	25 (56.8)	0.002	2.63	1.09-6.44	0.032
Kidney disease	8 (12.9)	10 (22.7)	0.184
Neuropathy	7 (11.3)	8 (18.2)	0.316
Cardiovascular disease	0	1 (2.3)	0.415
Peripheral arteriopathy	1 (1.6)	2 (4.5)	0.569
HbA1c (%)	7.6 ± 0.9	8.1 ± 1.1	0.017	1.83	1.15-3.06	0.015
Gastrointestinal symptoms	27 (43.5)	13 (29.5)	0.143
BMI (kg/m2)	24.7 ± 3.9	26 ± 4.9	0.183
Serum nutritional markers
Hemoglobin (g/dL)	13.9 ± 1.3	14.3 ± 1.2	0.124
Mean corpuscular volume (fL)	92.2 ± 6.5	93.1 ± 5.8	0.449
Lymphocytes (× 106/L)	1713 ± 566	1920 ± 578	0.068
Prothrombin index	98.1 ± 4.1	98.6 ± 4.0	0.491
Total proteins (g/L)	7.0 ± 0.4	6.9 ± 0.4	0.507
Albumin (g/dL)	4.4 ± 0.3	4.4 ± 0.3	0.441
Prealbumin (mg/dL)	20.9 ± 5.4	20.8 ± 3.6	0.880
Retinol binding protein (mg/dL)	3.2 ± 1.3	3.2 ± 1.0	0.864
Transferrin (mg/dL)	232.9 ± 42.2	232.8 ± 42.9	0.995
Ferritin (µg/L)	100.5 ± 96.8	79.2 ± 71.2	0.226
Cholesterol (mg/dL)	180 ± 34.5	182 ± 33	0.798
Triglycerides (mg/dL)	64.3 ± 26.9	62.9 ± 40.4	0.220
Vitamin A (mg/L)	0.4 ± 0.2	0.4 ± 0.1	0.913
Vitamin D (ng/mL)	26.6 ± 10.3	23.7 ± 9.6	0.161
Vitamin E (mg/L)	13.3 ± 3.0	13.6 ± 3.9	0.642
Vitamin B12 (ng/L)	504.5 ± 286	367.5 ± 183	0.006
Folic acid (µg/L)	11.4 ± 5.2	11.7 ± 4.1	0.790
Calcium (mg/L)	9.4 ± 0.4	9.6 ± 0.3	0.027
Iron (mg/dL)	79.6 ± 31.5	91.1 ± 37.5	0.093
Magnesium (mg/dL)	1.9 ± 0.2	1.9 ± 0.2	0.180
Zinc (µg/dL)	75.3 ± 14.5	76.3 ± 13.5	0.742

Data were showed as mean ± SD or n (%). DM: Diabetes mellitus; FEC: Fecal elastase concentration; HbA1c: Glycated hemoglobin; BMI: Body mass index; OR: Odds ratio; 95%CI: 95% confidence interval.

No significant differences were observed in gastrointestinal symptoms nor in BMI. Most serum nutritional markers did not differ between groups, except for vitamin B12, which was significantly lower in the FEC-deficient group, and calcium, which was slightly higher. Further analyses showed no significant differences in the vitamin B12 Levels of participants with or without neuropathy (P = 0.343) and no correlations with diabetes duration (P = 0.228) nor with patient age (P = 0.210). Notably, abnormal vitamin B12 levels were observed in only 7 patients (1 with normal FEC and 6 with abnormal FEC), while slightly low calcium levels were detected in just 1 patient from the normal FEC group.

Follow-up analysis

A total of 66 patients participated in the second phase of the study, including 29 men (43.9%), with a mean age of 55.4 ± 11.3 years. Comparison of participants who continued vs those who did not revealed no significant differences in baseline characteristics (Supplementary Table 2).

Table 4 summarizes the clinical and biochemical characteristics at both time points. As expected, mean age increased over the follow-up period. Lifestyle factors showed a decrease in smoking prevalence from 26% to 18%, although this was not statistically significant, while alcohol consumption significantly decreased from 21% to 11% (P = 0.016).

Table 4 Clinical and laboratory characteristics of the patients at both timepoints.

Characteristics	Timepoint 1, n = 66	Timepoint 2, n = 66	P value
Age (year)	47 ± 11.2	55.4 ± 11.3	< 0.001
Male sex	29 (44)	29 (44)	1
Female sex	37 (56)	37 (56)
Smoking	17 (25.8)	12 (18.2)	0.125
Alcohol	14 (21.2)	7 (10.6)	0.016
DM duration in years	19.3 ± 9.7	27.6 ± 9.6	< 0.001
DM-related complications	31 (47)	41 (62.1)	0.006
Retinopathy	29 (43.9)	33 (50)	0.344
Kidney disease	12 (18.2)	15 (22.7)	0.375
Neuropathy	9 (13.6)	14 (21.2)	0.063
Cardiovascular disease	1 (1.5)	2 (3)	1
Peripheral arteriopathy	1 (1.5)	4 (6.1)	0.250
HbA1c (%)	7.7 ± 1	7.5 ± 0.8	0.020
Fecal elastase (µg/g), median (IQR)	227.5 (253)	171.5 (246)	0.094
Fecal elastase group (µg/g)			0.035
> 200	39 (59.1)	27 (40.9)
100-200	13 (19.7)	20 (30.3)
< 100	14 (21.2)	19 (28.8)
Gastrointestinal symptoms	25 (37.9)	17 (25.8)	0.077
Dyspepsia	21 (31.8)	15 (22.7)	0.180
Diarrhea	4 (6.1)	3 (4.5)	1
Abdominal pain	3 (4.5)	2 (3)	1
BMI (kg/m2)	25.4 (4.4)	25.8 (4.6)	0.110
Nutritional markers below LLN1
Prealbumin	34 (52.3)	35 (53.8)	0.835
Retinol binding protein	34 (52.3)	20 (31.7)	0.003
Vitamin A	7 (10.8)	2 (3.1)	0.125
Vitamin D	21 (32.8)	37 (56.9)	0.007

¹Only parameters with a frequency of ≥ 10% cases in any of the time points are displayed in the table.

Data were showed as mean ± SD or n (%). DM: Diabetes mellitus; HbA1c: Glycated hemoglobin; BMI: Body mass index; IQR: Interquartile range (Q1-Q3); LLN: Lower limit of normal.

The duration of diabetes naturally increased during follow-up (P < 0.001). Despite a slight but significant reduction in HbA1c levels from 7.7% to 7.5% (P = 0.020), the overall prevalence of diabetes-related complications rose from 47% to 62.1% (P = 0.006). However, changes in the prevalence of individual complications were not statistically significant.

Changes in FEC over time

FECs showed a decreasing trend over the follow-up period, with median values declining from 227.5 µg/g to 171.5 µg/g; however, this change was not statistically significant. Categorizing participants by FEC revealed a decline in the proportion with normal levels (from 59.1% to 40.9%), while those with moderate and severe deficiency increased (19.7% to 30.3% and 21.2% to 28.8%, respectively), resulting in an overall significant shift in distribution (P = 0.035; Figure 1). No clinical nor demographic factors were identified as predictors of progression from normal to pathological FEC.

Figure 1 Evolution of fecal elastase deficiency.

Comparison of clinical and nutritional characteristics between time points

Gastrointestinal symptoms were reported by 38% of participants at the first assessment and decreased to 26% at follow-up, though this change was not statistically significant (P = 0.077). The prevalence of specific symptoms, including diarrhea, abdominal pain, and dyspepsia, remained largely unchanged.

Serum nutritional deficiencies were uncommon and primarily involved prealbumin, RBP, and vitamins A and D. Over the follow-up period, the prevalence of RBP deficiency significantly decreased, whereas vitamin D deficiency increased. BMI showed a slight, non-significant increase from 25.4 ± 4.4 kg/m² to 25.8 ± 4.6 kg/m².

DISCUSSION

Our results underscore the complex and incompletely understood relationship between the endocrine and exocrine pancreas, specifically highlighting impairment of exocrine function in T1D. At baseline, 41% of participants had pathologically low fecal elastase levels, increasing to 58% by the end of follow-up. These findings align with previous reports, which have documented prevalence rates between 16% and 57%[5,6,11-16].

In our cohort, fecal elastase deficiency was initially associated with male sex, higher HbA1c level, and the presence of diabetes-related complications, particularly retinopathy. There is currently no consensus in the literature regarding factors linked to low FEC in T1D. Some studies have reported associations with male sex and older age[11], but most have not confirmed these findings. Similarly, the findings on the relationship between exocrine pancreatic dysfunction and variables such as diabetes duration[11-13,17], glycemic control (HbA1c)[11,16], and the presence of diabetes-related complications[7] remain inconsistent. As with the prior reports, we found no association between alcohol and tobacco consumption and exocrine dysfunction.

It remains debated whether these changes reflect disturbed islet function or concurrent damage to both endocrine and exocrine pancreatic tissues. Nevertheless, low FEC did not appear to have meaningful clinical consequences in our cohort, showing no impact on BMI, gastrointestinal symptoms, nor nutritional status. The relationship between reduced FEC and low BMI in T1D remains unclear; only two studies have reported a weak association[12,16], while most others have found no link. Regarding gastrointestinal symptoms, only one study identified an association between FEC < 100 µg/g and diarrhea[11]. A small subset of patients (6 with low FEC and 1 with normal FEC) exhibited vitamin B12 deficiency. Low vitamin B12 levels have been reported previously in individuals with T1D and may result from malabsorption due to concomitant autoimmune conditions, such as celiac disease or autoimmune gastritis[18]. In addition, pancreatic proteases are known to play a role in the binding of vitamin B12 to intrinsic factor, facilitating its absorption in the distal ileum; however, vitamin B12 deficiency remains uncommon in patients with EPI. Notably, this deficiency showed no association with either neuropathy or patient age or diabetes duration.

This apparent lack of clinical relevance raises question about labeling low FEC in T1D as EPI. Notably, only 2 patients with low FEC received pancreatic enzyme replacement therapy: 1 following pancreatic surgery for a neuroendocrine tumor (excluded from follow-up) and 1 after the second FEC measurement due to concurrent diarrhea. In the latter case, endoscopic ultrasound revealed six criteria consistent with chronic pancreatitis, and the patient had a history of heavy smoking and alcohol consumption.

A possible explanation for the limited clinical impact may involve selective impairment of elastase-1 secretion, which could account for the poor concordance of FEC with the findings from the mixed ¹³C-triglyceride breath test[8,17] and secretin-cerulein test[7] in patients with T1D. Indeed, some authors have suggested that functional decline of different pancreatic enzymes may not occur simultaneously[19]. Regardless, further evaluation may be warranted in patients presenting with concurrent gastrointestinal symptoms or established risk factors for chronic pancreatitis.

Our results suggest that fecal elastase deficiency may progress over time, although the changes observed in our study’s population did not reach statistical significance. Similarly, Dozio et al[20] reported significantly lower FEC in patients with long-standing T1D compared to newly diagnosed patients, whose levels were comparable to healthy controls. They proposed parallel pathways affecting both endocrine and exocrine pancreatic function. However, in our cohort, the factors driving these changes remain unclear. Aging itself has been associated with a gradual decline in fecal elastase levels[21], and that could also have contributed to the observed findings.

Interestingly, exocrine function normalized in 4 of 27 patients with initial elastase deficiency. Marked intra-patient variability in FEC has been reported in cystic fibrosis, with coefficients of variation reaching up to 35%[22], but this phenomenon has not been investigated in T1D. Bidirectional changes in exocrine function were also assessed by Creutzfeldt et al[9] in 20 patients with T1D using the secretin-pancreozymin test. Among the 13 patients who initially showed normal exocrine function, 2 developed dysfunction, while 2 of 7 patients with baseline dysfunction showed normalization. Consistent with our findings, these changes occurred independent of diabetes duration, though other contributing factors were not assessed.

Changes during follow-up were not limited to FEC. Significant variations were also observed in vitamin D levels, with an increasing number of patients showing deficiency, and in RBP levels, which changed in the opposite direction. The methods used to measure these substances remained consistent throughout the study, so the reasons for these variations are unclear. Although vitamin D is known to fluctuate seasonally, this factor was not analyzed in our sample. It is also possible that aging contributed to the progression toward vitamin D deficiency, as advancing age is a well-established factor associated with reduced vitamin D status. Vitamin D deficiency in T1D is likely multifactorial, involving reduced sun exposure, limited dietary intake, impaired absorption, and genetic factors affecting vitamin D metabolism. Notably, vitamin D deficiency was not associated with FEC status in either the first or second phase of the study. Regarding RBP, the observed changes remain unexplained, but considering that the observed vitamin A deficiency was rare and did not differ between the study phases, its clinical relevance may be limited.

Prealbumin and RBP deficiencies were common, consistent with findings from our previous work[23] and other studies[20,24,25]. As in our prior study, the levels of these proteins were strongly correlated, with significantly lower concentrations observed in women (data not shown). For prealbumin, a potential explanation may involve changes in its quaternary structure. In T1D, the tetrameric form decreases while the monomeric form increases. Current assays primarily measure the tetrameric form. This structural change may have pathophysiological relevance in T1D, as the tetramer enhances glucose-stimulated insulin secretion and protects pancreatic beta cells from apoptosis, whereas the monomer lacks these functions[26].

Our study has several limitations. Forty participants (37.7% of the initial cohort) did not complete the second phase, with thirty of those having declined to participate in the second phase. There may be several reasons underlying this profile of refusal for study continuance. First, patients with T1D are often diagnosed at an early age and consequently have a long history of intensive medical follow-up, including frequent clinic visits, regular consultations with multiple specialists, and numerous laboratory and imaging tests. This prolonged clinical burden may reduce willingness to participate in additional research studies. Furthermore, the collection of stool samples, required for the determination of fecal elastase, is plagued by a poor acceptance rate even in the general population. Nevertheless, the final cohort included more than three times the number of patients compared to a previous study[9] and, to our knowledge, this is the first study to use FEC to assess longitudinal changes in exocrine pancreatic function in individuals with T1D. Importantly, patients who completed follow-up and those who dropped out had similar baseline characteristics.

It is well established that the positive predictive value of FEC is limited in populations with a low pretest probability of EPI[27]. Accordingly, another limitation of this study is the lack of a gold standard measure of exocrine pancreatic function, such as the coefficient of fat absorption. However, while highly accurate, the cumbersome and time-consuming nature of that test make it impractical for routine clinical use and support the continued preference for FEC measurement[28]. Nevertheless, future studies incorporating direct tests in T1D patients are warranted to better define the prevalence and clinical significance of exocrine pancreatic dysfunction. Finally, we were unable to identify factors associated with FEC decline over time. Previous reports on predictors of low FEC are heterogeneous, likely reflecting the complexity of the test, which actually measures chymotrypsin-like elastase 3A and 3B, and its inherent limitations, particularly in conditions other than chronic pancreatitis[19].

CONCLUSION

Overall, our findings highlight the natural progression of diabetes and its complications over time, along with associated changes in exocrine pancreatic function, emphasizing the importance of evaluating their clinical relevance for patient management. In summary, exocrine pancreatic function appears to be altered in T1D and may deteriorate over time; however, the underlying pathophysiology remains unclear. Nevertheless, its clinical impact seems limited, so low fecal elastase levels in T1D should be interpreted with caution to avoid overdiagnosis and unnecessary treatments.