Pancreatic Enzyme Replacement Leads to Increased Vitamin D Uptake in Patients Undergoing Sleeve-gastrectomy — A Prospective, Monocentric Trial

Abstract

Purpose

Metabolic and bariatric surgery (MBS) is often considered to be associated with macro- and micronutrient deficiency. A possible treatment option can be the implementation of pancreatic enzyme replacement (PERT) and may lead to better outcomes. We designed a prospective trial investigating the possible impact of PERT in patients undergoing MBS at a high-volume center.

Materials and Methods

A prospective two-arm randomized controlled trial was conducted on patients who underwent either sleeve gastrectomy or gastric bypass procedures at a high-volume center. Patients underwent bariatric surgery and follow-up examinations at 3, 6, and 12 months after surgery. Patients were stratified either to the treatment group with PERT or to the control group. The primary endpoint of the study was a change in BMI. Lab testing was carried out to measure secondary endpoints, including albumin and vitamin D levels.

Results

Overall, 204 patients were enrolled. Due to missing follow-ups, surgical complications, and side effects due to Kreon medication, 65 were excluded. Analysis of primary endpoints indicates that PERT does not lead to slower weight loss or BMI reduction. Analysis of secondary endpoints showed significantly better vitamin D levels in patients undergoing MBS and PERT. No statistical difference was seen regarding albumin. In both arms, fatty liver disease improved. Quality of life is positively judged as comparable by patients in both groups.

Conclusion

Herein, we show an association between PERT and higher vitamin D levels in patients undergoing MBS. An optimized enzymatic environment due to PERT may therefore result in higher vitamin D levels and may improve clinical outcomes in patients undergoing MBS.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s11695-024-07526-5.

Introduction

Extreme obesity has increased over the past decades across the world and is associated with significant morbidity and mortality [1]. Obesity is associated with severe co-morbidities such as insulin resistance and diabetes, cardiovascular diseases, non-alcoholic fatty liver disease, musculoskeletal disorders, and many more including cancer [1–3]. Until today, bariatric surgery (MBS) remains the most efficient treatment option in regard to weight loss and control of metabolic disorders also in the long term. The basic principle of MBS is restriction and malabsorption due to surgical introduced anatomical changes primarily of the upper gastrointestinal tract. MBS aims at upper digestive tract modification including restriction in size mainly with a focus on the stomach with or without implementing bypass procedures of the small intestines [4–6].

Due to modifications of the GI tract, bariatric surgery is often considered to be associated with macro- and micronutrient deficiency. Also, implementing restriction and malabsorption is commonly associated with gastrointestinal symptoms such as diarrhea, bloating, abdominal pain, and nausea and can lead to significant morbidity in patients undergoing bariatric surgery [7]. Recently, exocrine pancreatic insufficiency (EPI) and more specifically pancreaticocibiale asynchrony are becoming increasingly recognized as contributors to gastrointestinal symptoms after MBS. But incidence and outcome still remain controversial, and the underlying mechanisms are not well understood. A possible treatment option can be the implementation of pancreatic enzyme replacement (PERT) with food intake which may lead to symptom control and improved outcome. A recent study analyzed retrospectively possible impact and found EPI in over 40% of patients accounting for GI symptoms, and PERT improved symptoms in almost 87% of them [8]. Taken together, untreated EPI results in substantial patient harm including severe morbidity including steatorrhea and nausea and impacts protein and fat metabolism. However, routine application of PERT after MBS is not standard.

Therefore, we designed a prospective trial investigating the possible impact of PERT in patients undergoing MBS at a high-volume center.

Methods

Patient Population

A prospective two-arm randomized controlled trial was conducted on patients who underwent either sleeve gastrectomy or gastric bypass procedures between 2022 and 2023 at a high-volume center. Due to only a small number of patients undergoing bypass surgery, patients receiving SG were further analyzed only. Patients were recruited during routine obesity consultation hours and included when they met the following criteria: age 18 or older, grade II obesity and metabolic syndrome, respectively, and patients with grade III obesity with or without metabolic syndrome. Patients underwent bariatric surgery and follow-up examinations at 3, 6, and 12 months after surgery. Exclusion criteria were pregnancy, age under 18, not being able to give consent, or when patients did not fulfill the guidelines of the German Adipositas Society, and the German Society for General and Visceral Surgery, respectively, for surgical therapy.

Ethics

The study was approved by the relevant Ethics Committee Westfalen-Lippe, Germany, on December 23^rd, 2021 (Nr. 2021-243-f-S). All research procedures were conducted in accordance with the Declaration of Helsinki and the professional regulations of the medical association Westfalen-Lippe in Germany.

Study Design

Patients were stratified either to the treatment group with pancreatic enzyme substitution (PERT group) or to the control group. PERT was carried out by Kreon^® application (pankreatin; mix of pancreatic enzymes) as used standardized in the treatment of EPI. Pancreatic enzyme replacement using Kreon^® (pankreatin) manufacturer’s specifications (www.kreon.de). As defined by the study protocol, 25.000 units (IE) were given just before the meal. Randomization was carried out randomly using a block-randomization scheme with blocks of size four and balanced randomization into the treatment and control arm. Details of the study design are shown in the flow chart (Fig. 1). Prior to surgery, patients underwent a pre-operative body composition assessment, including weight, height, BMI, fat mass (FM), waist size, and phase angle measurements. Body composition was assessed using the bioelectrical impedance analysis (seca B.I.A. scale). B.I.A. analysis is the gold standard until today for the assessment of body composition as described elsewhere independently of weight and BMI (www.seca.com). The primary endpoint of the study was the change in BMI, and measurements were undertaken at 3, 6, and 12 months after surgery. Lab testing was carried out to measure secondary endpoints, including albumin and vitamin D levels since obese patients often present with vitamin D deficiency and malnutrition including for protein intake and quality.

Fig. 1.

Study design of the proposed trial: 204 patients were randomized with a block randomization scheme and blocks of size 4 equally into treatment and control; several surgical procedures were applied, the choice of which was among the discretion of the surgeon

Also, the degree of fatty liver disease was recorded with ultrasound examinations as further secondary endpoints. In particular, liver morphology was assessed by ultrasound as a secondary endpoint indicating loss of fat accumulation in the liver after MBS. Possible changes regarding the liver function were not assessed and were not an endpoint of the study. Two examinators used a Hitachi Arietta V70 ultrasound system during the follow-up. As a standardized approach in clinical routine, the hepatorenal index was measured for each patient at baseline and after 6 and 12 months postoperatively. The presence of cholecystolithiasis was documented in the same way [9, 10].

All patients undergoing MBS received standardized supplementation regarding protein, vitamins, and trace elements as recommended by the German Nutrition Society (DGE). Regular supplementation was assessed during routine follow-up as described. Details are shown in Table S2 (see Table S2 in the supplemental file).

Additionally, quality of life was measured with the standardized bariatric quality of life (BQL) questionnaire. Comorbidities and medication as well as complications during surgical intervention were recorded.

Statistical Analysis

The primary endpoint of the study was BMI, proposing that PERT should not affect BMI respectively decrease weight loss. With regard to the secondary endpoint as shown above, vitamin D levels were analyzed, since decreased vitamin D levels are common in bariatric patients, and vitamin D is associated with metabolic outcomes as shown above. Baseline data in both groups were assessed exploratory and with the Mann-Whitney U test for BMI and Welch’s two-sample two-sided t-test for weight, fat mass, waist size, and phase angle at baseline, to detect significant differences which undermine the effect of randomization between treatment and control group. Vitamin D and albumin were assessed at baseline with Welch’s two-sample two-sided t-test for the same reason. Statistical significance was established at $p<0.05$, with Bonferroni correction for multiple testing. The analysis at 12 months post-surgery for BMI was conducted with the Mann-Whitney U test, where sample size calculations yielded 100 patients per group to detect effect sizes up to 1866 kg/m² with 80% power. This equals a standardized effect size of Cohen’s $=0.4$ , when significance is established at $p<0.05$, and multiple testing for BMI and vitamin D level is accounted for by applying the Benjamini-Hochberg correction. Thus, the total trial sample size was planned at 200 patients, 100 in each trial arm. A block-randomization scheme was used to randomize patients equally in both trial arms, with block sizes of four patients.

Results

A total of 204 patients were enrolled in the study. Only 21 patients underwent gastric bypass, the majority of which did miss follow-up examinations or dropped out because of intolerance to PERT medication. As a consequence, these cases were removed, and all remaining patients analyzed underwent laparoscopic sleeve resection. Due to missing follow-ups, surgical complications, and/or side effects due to pancreatic enzyme replacement, patients were further reduced to n=139. Patient characteristics are shown in Tables 1 and 2; the mean age was 42.6 years, the population was predominantly female (67.62%), and the mean BMI at baseline was 49.35 kg/m².

Table 1.

Descriptive statistics for primary endpoints baseline; Kreon^®: pancreatic enzyme replacement medication

	Age baseline		Height baseline		BMI baseline		Weight baseline		Fat mass baseline		Waist size baseline		Phase angle baseline
	Control	Kreon	Control	Kreon	Control	Kreon	Control	Kreon	Control	Kreon	Control	Kreon	Control	Kreon
Valid	79	60	79	60	79	60	79	60	78	60	78	60	78	60
Missing	0	0	0	0	0	0	0	0	1	0	1	0	1	0
Median	39.000	44.500	169.000	173.000	50.460	48.080	138.600	144.975	71.765	71.335	135.500	134.500	5.400	5.150
Mean	40.456	45.567	168.266	173.733	49.914	48.712	141.803	147.410	73.386	73.488	134.013	134.800	5.373	5.090
Std. deviation	10.328	12.037	8.553	11.132	6.586	6.566	24.744	26.638	14.902	15.508	13.900	15.824	0.706	0.651
IQR	13.500	19.500	13.000	13.250	7.010	8.505	31.200	35.338	16.067	20.365	18.250	19.500	0.700	1.000
Minimum	23.000	24.000	151.000	150.000	35.290	37.490	96.550	107.350	42.850	48.610	100.000	107.000	3.100	3.500
Maximum	61.000	70.000	188.000	198.000	65.900	66.120	215.900	226.700	123.380	107.260	164.000	169.000	7.300	6.200

Table 2.

Frequencies for gender baseline

Treatment	Gender baseline	Frequency	Percent	Valid percent	Cumulative percent
Control	M	21	26.582	26.582	26.582
	W	58	73.418	73.418	100.000
	Missing	0	0.000
	Total	79	100.000
Kreon	M	24	40.000	40.000	40.000
	W	36	60.000	60.000	100.000
	Missing	0	0.000
	Total	60	100.000

No significant differences in BMI, FM, weight, and waist size were found at baseline (Table 3 and Fig. 2). However, despite randomization, the control group had significantly larger phase angle values at baseline, compared to Table 1. This is also highlighted in the lower left boxplot in Fig. 2. Table 4 shows that there were no significant differences in albumin or vitamin D levels. This is important, because due to this fact, significant differences at 12 months post-surgery in vitamin D or albumin levels can be attributed to the application of pancreatic enzyme replacement and are not due to already apparent systematic differences between treatment and control at baseline.

Table 3.

Independent samples Welch’s t-tests for baseline differences of weight, fat mass, waist size, and phase angle

	t	df	p	Mean difference	SE difference	Cohen’s d	SE Cohen’s d	95% CI for Cohen’s d
								Lower	Upper
Weight baseline	−1.267	122.028	0.896	−5.607	4.424	−0.218	0.172	−0.500	∞
Fat mass baseline	−0.039	124.464	0.515	−0.102	2.618	−0.007	0.172	−0.289	∞
Waist size baseline	−0.305	117.976	0.620	−0.787	2.579	−0.053	0.172	−0.335	∞
Phase angle baseline	2.439	131.570	0.008	0.283	0.116	0.417	0.175	0.130	∞

For all tests, the alternative hypothesis specifies that group Control is greater than PERT group

Welch’s t-test

Fig. 2.

Boxplots for BMI, weight, fat mass, phase angle, waist size, and age at baseline for control and Kreon groups

Table 4.

Independent samples Welch’s t-tests for secondary endpoints vitamin D and albumin levels at baseline

	t	df	p	Mean difference	SE difference	Cohen’s d	SE Cohen’s d	95% CI for Cohen’s d
								Lower	Upper
Vitamin D baseline	−1.042	67.996	0.301	−2.061	1.978	−0.249	0.241	−0.719	0.223
Albumin baseline	−1.462	117.004	0.147	−0.072	0.050	−0.257	0.176	−0.602	0.088

Welch’s t-test

Twelve months postoperatively, there was a mean BMI of 36.83 kg/m² (SD=6.52) in the control group and 36.33 kg/m² (SD=7.08) in the PERT group. As shown in Table 5, fat mass, waist size, and weight are comparable between PERT and the control group 12 months after surgery. Figure 3 shows the distribution of BMI, weight, fat mass, phase angle, and waist size 12 months post-surgery. Both groups achieved a statistically significant reduction of BMI with no apparent changes between the PERT and control group, indicated by a nonsignificant Mann-Whitney U statistic with a p-value of $p=0.502$, compared to Table 6. Weight, fat mass, and waist size all reduced significantly in both groups with no apparent differences between PERT and control arm, compared to Tables 1 and 5. While patients in the control and PERT arms at baseline had mean BMI of 49.91 ($SD=6.59)$ and 48.71 $(SD=6.57)$, at 12 months after surgery, these values reduced to 36.83 ($(SD=6.53)$ and 36.34 $(SD=7.08)$.

Table 5.

Descriptive statistics 12 months post-surgery

	BMI 12 months		Weight 12 months		Fat mass 12 months		Waist size 12 months		Phase angle 12 months
	Control	Kreon	Control	Kreon	Control	Kreon	Control	Kreon	Control	Kreon
Valid	79	59	79	59	77	58	77	58	77	58
Missing	0	1	0	1	2	2	2	2	2	2
Median	36.310	34.570	102.500	102.900	44.460	43.040	102.000	102.500	4.700	4.500
Mean	36.831	36.338	105.033	109.768	43.814	43.654	100.678	104.382	4.827	4.519
Std. deviation	6.526	7.082	23.970	24.681	14.133	15.492	23.510	23.022	0.766	0.826
IQR	8.100	8.725	30.475	35.750	18.550	18.565	27.000	30.500	0.900	0.950
Minimum	24.200	23.680	55.950	56.900	17.240	14.960	0.990	1.130	2.800	0.800
Maximum	55.680	54.170	182.400	154.400	86.290	78.400	140.000	149.000	7.100	6.500

Fig. 3.

Boxplots for BMI, weight, fat mass, phase angle, and waist size 12 months after surgery

Table 6.

Mann-Whitney-U-test for BMI 12 months post-surgery

	W	df	p	Hodges-Lehmann estimate	Rank-biserial correlation	SE rank-biserial correlation	95% CI for rank-biserial correlation
							Lower	Upper
BMI 12 months	2487.000		0.502	0.860	0.067	0.099	−0.127	0.257

For the Mann-Whitney test, the effect size is given by the rank biserial correlation

Phase angle measurements were carried out to assess the nutritional status of the body cells as described elsewhere further indirectly indicating body health and nutritional status. Phase angle measurements remained significantly different between both groups at 12 months postoperatively, compared to Table 5 and Fig. 2, with a decrease compared to baseline, both in the PERT and control group.

In summary, analysis of primary endpoints indicates that PERT with pancreatic enzyme replacement does not lead to slower weight loss or BMI reduction 12 months after bariatric surgery. Furthermore, BMI reduction and weight loss were comparable in both trial arms.

The analysis of secondary endpoints showed significant differences in vitamin D levels between the PERT and control group (Fig. 4), as tested with Welch’s two-sample t-test (one-sided) (Table 7), yielding $t=-1.78$, with 129.34 degrees of freedom, $p=0.038$, and Cohen’s $d=-0.308(SD=0.175)$, also shown in Table 9. Thus, vitamin D levels improved significantly in the PERT group, showing a moderate clinical effect (Table 8). Before MBS, only ten patients regardless of the study group took vitamin D prior to surgery without any statistical difference at baseline (data not shown). Albumin levels increased, but no statistically significant differences between both groups could be found, comparing Table 9 and Fig. 4. We can only speculate whether a larger sample size would reveal a significant effect, which can ultimately only be answered by further clinical trials tackling PERT in patients after bariatric surgery.

Fig. 4.

Vitamin D and albumin levels 12 months after bariatric surgery

Table 7.

Welch’s two-sample independent t-tests for weight, fat mass, waist size, and phase angle at 12 months after surgery

	t	df	p	Mean difference	SE difference	Cohen’s d	SE Cohen’s d	95% CI for Cohen’s d
								Lower	Upper
Weight 12 months	−1.129	123.079	0.869	−4.735	4.195	−0.195	0.173	−0.478	∞
Fat mass 12 months	0.062	116.519	0.475	0.160	2.595	0.011	0.174	−0.275	∞
Waist size 12 months	−0.917	124.228	0.819	−3.703	4.039	−0.159	0.174	−0.445	∞
Phase angle 12 months	2.215	117.753	0.014	0.308	0.139	0.387	0.177	0.097	∞

For all tests, the alternative hypothesis specifies that group Control is greater than the PERT group

Welch’s t-test

Table 9.

Independent sample T-test for secondary endpoints vitamin D and albumin 12 months after bariatric surgery

	t	df	p	Mean difference	SE difference	Cohen’s d	SE Cohen’s d	95% CI for Cohen’s d
								Lower	Upper
Vitamin D 12 months	−1.786	129.343	0.038	−4.063	2.275	−0.308	0.175	−∞	−0.020
Albumin 12 months	−1.473	100.757	0.072	−0.077	0.052	−0.289	0.199	−∞	0.038

For all tests, the alternative hypothesis specifies that group Control is less than the PERT group

Welch’s t-test

Table 8.

Descriptive statistics for secondary endpoints 12 months after bariatric surgery

	Vitamin D 12 months		Albumin 12 months
	Control	Kreon	Control	Kreon
Valid	76	59	58	46
Missing	3	1	21	14
Median	32.750	34.800	4.345	4.441
Mean	33.200	37.263	4.379	4.456
Std. deviation	13.748	12.595	0.282	0.250
IQR	16.325	12.550	0.280	0.357
Minimum	5.200	13.900	3.770	3.920
Maximum	76.300	79.400	5.030	4.950

Analysis of fatty liver disease showed no statistical differences between both trial arms at baseline (Fig. S1 in the supplemental file). In both arms, fatty liver disease improved postoperatively, with no conceivable differences between the PERT and control group (Fig. S2 in the supplemental file).

Importantly, the BQL questionnaire handed to patients at 12 months after surgery assessed their current quality of life. As shown below (Fig. 5), quality of life is positively judged comparable by patients taking pancreatic enzymes and controls, indicating that improved vitamin levels and comparable BMI reduction when administering PERT with pancreatic enzyme replacement leads to no apparent changes in the perceived quality of life in patients. Whether the improved vitamin D levels are causal to the slight increase in percentages of the categories “very good” and “rather good” cannot be fully answered by this trial.

Fig. 5.

Answers to the bariatric quality of life questionnaire item “How do you judge your current quality of life?” of patients 12 months after bariatric surgery for groups Kreon (right pie chart) and control (left pie chart)

Discussion

Several surgical procedures have been introduced in the past decades, but until today, laparoscopic sleeve resection (SG) and the laparoscopic Roux-en-Y gastric bypass remain the most widely performed procedures. In SG, approximately 80% of the stomach is removed along a digestive tube resulting in a tubular curved stomach body resulting in less food intake, impaired ghrelin response with reduced hunger, and improved metabolic outcome [4].

MBS affects the delicate physiology of pancreatic enzyme function and may result in EPI with disruptive symptoms and may further result in micro- and macronutrient changes including protein and vitamin intake. In our study, endpoint analysis did not aim to investigate possible pancreatic enzyme insufficiency prior to MBS or 12 months after MBS as described above. However, to our knowledge, no specific data are available on how SG possibly affects pancreatic enzyme function or if patients with severe obesity suffer under relevant EPI prior to surgery. Also, data known from patients undergoing non-bariatric upper GI procedures or undergoing pancreatic surgery are not comparable with patients undergoing MBS. Since general assessment of EPI was not the aim of the study, we did not assess specifically EPI using stool testing for elastase or other testing assessing pancreatic enzyme function. Vitamin D deficiency is typically seen in patients with obesity and inversely correlates with BMI and obesity [11, 12]. Vitamin D affects insulin action and glucose metabolism via glucose transportation in adipocytes and suppresses various inflammatory cytokines resulting in reduced inflammation. Therefore, vitamin D is strongly influential on metabolic status and disorders [13–15]. Actual guidelines on vitamin D status mainly focus on bone disorders like osteoporosis or osteomalacia as well as on kidney diseases [16]. More recently, the bidirectional relationship between vitamin D metabolism and obesity as well as its impact on vitamin D levels in patients undergoing bariatric surgery was emphasized. Patients undergoing bariatric surgery typically appear to have low vitamin D levels, which may result in even lower levels after bariatric procedures. Also, bariatric surgery is associated with a postoperative reduction of vitamin D levels particularly after malabsorptive procedures, and life-long supplementation is recommended after bariatric surgery [17].

Herein, for the first time, we show an association between PERT and higher vitamin D levels in patients undergoing SG exclusively. As the association between metabolic outcome and vitamin D levels is well established [12, 14, 15], PERT may improve the intrinsic effects of weight loss and metabolic control in patients undergoing sleeve gastrectomy. Pathophysiological vitamin D absorption and metabolism may be disturbed in disorders that cause poor fat absorption resulting in limited vitamin D absorption. Besides others, MBS cause malabsorption as shown above, and an optimized enzymatic environment due to PERT may therefore result in higher vitamin D levels and may improve clinical impact at least in patients undergoing SG but also in patients undergoing other bariatric procedures [18]. However, we did not further explore the possible correlation between vitamin D levels and patient outcomes undergoing bypass surgery due to only a small patient cohort. Also, we did not explore a possible effect of increased vitamin D levels in patients exploring PERT after MBS on associated micronutrients including calcium to avoid multiplications. However, it could be interesting to test the impact of increased vitamin D levels on calcium metabolism in further studies. Furthermore, also, no correlation was found between weight loss and vitamin D levels as well as possible improvement in quality of life. However, we did recognize a positive trend seen at least in our study (see Table S2 in the supplemental file) also not being statistically significant, which may be due to the small patient cohort analyzed or insufficient follow-up time.

Taken together, further clinical trials may reveal a stronger correlation between clinical outcome and metabolic control once PERT is implemented routinely in supplementary diets not only for vitamin D levels in patients undergoing bariatric surgery.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent

Informed consent was obtained from all individual participants included in the study.

Conflict of Interest

The authors declare no conflict of interest.