*Review* **Sleeve Gastrectomy and Roux-En-Y Gastric Bypass. Two Sculptors of the Pancreatic Islet**

**Gonzalo-Martín Pérez-Arana 1,2,3,\*, José Fernández-Vivero 1, Alonso Camacho-Ramírez 1,3,4, Alfredo Díaz Gómez 5, José Bancalero de los Reyes 6, Antonio Ribelles-García 1, David Almorza-Gomar 2,7, Carmen Carrasco-Molinillo <sup>1</sup> and José-Arturo Prada-Oliveira 1,2,3,\***


**Citation:** Pérez-Arana, G.-M.; Fernández-Vivero, J.; Camacho-Ramírez, A.; Díaz Gómez, A.; Bancalero de los Reyes, J.; Ribelles-García, A.; Almorza-Gomar, D.; Carrasco-Molinillo, C.; Prada-Oliveira, J.-A. Sleeve Gastrectomy and Roux-En-Y Gastric Bypass. Two Sculptors of the Pancreatic Islet. *J. Clin. Med.* **2021**, *10*, 4217. https://doi.org/10.3390/ jcm10184217

Academic Editor: David Benaiges Boix

Received: 22 July 2021 Accepted: 14 September 2021 Published: 17 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Abstract:** Several surgical procedures are performed for the treatment of obesity. A main outcome of these procedures is the improvement of type 2 diabetes mellitus. Trying to explain this, gastrointestinal hormone levels and their effect on organs involved in carbohydrate metabolism, such as liver, gut, muscle or fat, have been studied intensively after bariatric surgery. These effects on endocrine-cell populations in the pancreas have been less well studied. We gathered the existing data on these pancreatic-cell populations after the two most common types of bariatric surgery, the sleeve gastrectomy (SG) and the roux-en-Y gastric bypass (RYGB), with the aim to explain the pathophysiological mechanisms underlying these surgeries and to improve their outcome.

**Keywords:** sleeve gastrectomy; roux-en-Y gastric bypass; beta-cell; alpha-cell; epsilon-cell; islet; trans-differentiation

### **1. Introduction**

Bariatric/metabolic surgery has been a powerful tool for the treatment of diabetes mellitus for a long time. Sleeve gastrectomy (SG) and roux-en-Y gastric bypass (RYGB) are two of the most performed ones [1,2] as Figure 1 shows.

Changes in energy homeostasis and body fat mass have been proposed as a primary mechanism to explain these phenomena [3,4], but other mechanisms such as changes in several gastrointestinal hormones also seem to be involved with a large number of publications written on the topic. Many of them have related the anatomical changes in the gastrointestinal tract after surgery with the modification of serum levels of glucagon like peptide-1 (GLP-1) [5], ghrelin [6], peptide tyrosine-tyrosine (PYY) [7], gastrointestinal inhibitory peptide (GIP) [8], or even leptin [9], among others, in humans and animal models. Their involvement is clear, but the exact mechanisms and their degree of participation remain partially unknown.

At the other end of the entero-pancreatic axis, the endocrine pancreas containing Langerhans islets determines changes in carbohydrate metabolism after bariatric/metabolic surgery. Their hormonal secretions before and after bariatric/metabolic surgery have been widely studied in plasma or serum from animals and humans [10,11] but the islet cell composition and its paracrine interactions have been studied less. We will attempt to

summarize what we know about the subject by means of a bibliographical review of the most relevant works published on the subject.

**Figure 1.** Schematic drawing of Sleeve Gastrectomy and Roux-en-Y Gastric bypass. (**A**) Sleeve Gastrectomy (SG). Representation of a common human sleeve gastrectomy (SG) procedure. The SG is a surgical procedure including a reduction of final gastric volume, since most of the gastric major curvature is resected. The stomach is reduced to a cylindrical pouch removing most of the fundus, stomach-corpus and antrum. The pylorus and minor curvature is preserved. SG reduces the initial stomach volume by approximately 15–20%. In animal models this configuration is maintained since the final gastric pouch volume and valves are preserved. (**B**) Roux-en-Y Gastric Bypass (RYGB). Representation of a common human roux-en-Y gastric bypass (RYGB) surgery. This includes a transverse section of the stomach performed from the major to the minor curvature, configuring a gastric pouch. This pouch of the stomach continues to the food handle with an alimentary bulb, which continues with the medium portion of the jejunum. RYGB, a mixed malabsorptive and restrictive technique, excludes the antrum and the proximal intestine to aliments by bypassing the duodenum and the initial part of the jejunum. This includes biliopancreatic secretion, which determines the malabsorptive component. The biliopancreatic bulb connects with the mid jejunum. In rats, the model was reproduced similarly with minor modifications according to the animal anatomy. Exempli gratia, the jejunal alimentary bulb was 10 cm due to the usual intestinal medium extension of 80 cm. Original figure seen in https://sagebariatric.com/about-surgery-home/sleeve-gastrectomy (accessed on 22 July 2021).

#### **2. Methods and Results**

This paper is a narrative literature review text that aims to expose the framework surrounding the effects of RYGB and SG on endocrine-cell populations in the pancreas. We performed a selective search of numerous articles in different databases, as well as books.

The literature of the main scientific databases was reviewed. The search was limited to documents published between 2001 and 2021. These databases were Medline, PubMed, Chochrane and Scopus. In addition, a search was carried out on academic websites, such as Google Scholar, SciELO and Dialnet. The main Boolean operators used were: AND, OR and NOT, and the key words were sleeve gastrectomy; roux-en-Y gastric bypass; beta-cell, alpha-cell; epsilon-cell; islet; trans-differentiation. Due to the large number of studies found, the following criteria were applied to filter the results and work with the most relevant studies.

Inclusion criteria: Original articles, systematic reviews and meta-analyses concerning modifications of the endocrine pancreas after bariatric or metabolic surgery in humans or animal models. Papers published in English in the last 20 years (2001–2021). We prioritised information from systematic reviews and meta-analyses with high scientific evidence. Exclusion criteria: Papers not related to the topic or not meeting the inclusion criteria. In the end, a total of 435 articles were found that met the search criteria. Of these, 47 were selected for the preparation of this manuscript. As Table 1 shows, a large number of disciplines are involved in the study of the topic.

**Table 1.** Search Results. Break down of the total number of articles used to prepare the work. The left column represents the different fields of research of each journal citation (Journal Citation Report categories). The central column contains the number of citations found in each category and the right column contains the number and percentage of citations selected for the manuscript.


#### **3. Discussion**

#### *3.1. The Sleeve Gastrectomy and the Islet Architecture*

Bariatric/metabolic surgery involves different techniques leading to different effects on pancreatic cell populations. Currently, sleeve gastrectomy (SG) is one of the most performed techniques. A consequence of this procedure is the drastic removal of the gastric fundus and corpus ghrelin-producing cell population. This situation leads to 35–45% reduction of blood ghrelin levels after gastrectomy in humans [12–14]. However, a recent study described the expansion of the pancreatic residual postnatal epsilon-cell population with recovery of plasma ghrelin levels in rats twelve weeks after SG. This expansion takes place at the expense of pancreatic cell progenitors that differentiate into epsilon-cells showing a high expression of lineage markers such as neurogenin-3 (Ngn-3) but not homeodomain protein Nkx2.2 (Figure 2) [15].

This leads us to believe in an adaptive response of the endocrine pancreas to low circulating ghrelin levels and in a possible explanation of the improvement of beta cell function after SG if we take into account the protective role of ghrelin on it [16].

Furthermore, this surgery does not only affect the epsilon-cells in the islets. It is clear that SG preserves the beta-cell function, at least for a while [17,18]. This could be explained by the increase of GLP-1 receptor expression in beta cells after SG, implying an increase in paracrine sensitivity to GLP-1 [19,20]. However, there are doubts about this due to a recent study with a modified mouse model involving an inducible knockdown of GLP-1r in beta-cells (GLP1rβ-cell-ko), which showed improved glycemic profiles, to the wild-nature level, after SG [21]. Other researchers have linked the maintenance of beta-cell mass and beta-cell identity markers such as PDX-1 or MafA [22] (Picture 2) to high levels of gastrin after SG, as well as to correction of long-term blood glucose levels in rodents [23].

**Figure 2.** Pancreatic endocrine cell identity markers and possible cell trans/differentiation pathways after SG/RYGB. Pancreatic endocrine-cell identity markers and possible cell differentiation pathways from progenitor-cells (Black arrows) or trans-differentiation from other pancreatic endocrine-cells (Blue arrows) after sleeve gastrectomy or roux-en-Y gastric bypass.

This brings us to the problem of diabetes relapse after SG, which is as high as 41.6% of cases five years after surgery [2]. Liu et al. proposed long-term recovery of insulin sensitivity without beta-cell dysfunction as an answer to the question [24], but a recent work showed loss of beta-cell mass and a strong increase in alpha-cell mass in Wistar rats twelve weeks after SG. Trans-differentiation of the beta-cell population under stressful situations with loss of beta-cell markers such as PDX-1 and gain of alpha-cell markers such as Pax-6 and Arx has been shown [25] (Figure 2). Moreover, this is supported by studies performed on mice outside the scope of bariatric surgery where alpha-cell populations labeled with Gcg-Cre lineage tracers showed a dilution of the marker at the expense of the beta-cell population throughout life [26]. Therefore, the appearance of alpha-cells at the expense of the beta-cell population may explain the long-term relapses in diabetes after SG.

Finally, the protective effect of the somatostatin-14 isoform on Min6 pancreatic beta cells of mice has recently been verified, limiting the stress markers HSPa1 and Ddit3 and apoptosis [27]. This together with the occurrence of delta-cell hyperplasia in Goto-Kakizaki diabetic mice [28] makes us think about a possible role of this delta population in the mechanisms underlying SG. This seems to be reinforced by the ability of ghrelin to activate the paracrine secretion of somatostatin [29] as mentioned above. However, due to the difficulty in carrying out these studies in humans and the ethical aspects, further investigation on animal models is needed to clarify this issue and the possible involvement of other pancreatic endocrine populations.

#### *3.2. The Roux-en-Y Gastric Bypass and the Islet Architecture*

Roux-en-Y gastric bypass appears to be the most powerful tool for the management of obesity and hyperglycemia in patients [30]. This procedure has demonstrated its efficiency in increasing beta-cell function in animal models and patients [31,32]. It also appears to increase beta-cell mass after surgery in both animal models and patients [33,34]. GLP-1 activity has been proposed as responsible for these effects on beta-cell mass after RYGB [35]. On the other hand, glucose improvement after RYGB has long been reported in mice models of functional GLP-1 and GLP-1 receptor deficiency, suggesting a GLP-1 independent mechanism for glycemic control after surgery [36]. Another very interesting candidate

is intra-islet PYY. Guida et al. reported a large increase in islet PYY content after RYGB, mediated by locally produced PYY but not GLP-1 glucose-stimulated insulin secretion. Furthermore, interleukin-22 (IL-22) seems to play a key role in the increase of intra-islet expression of PYY after RYGB. This situation would imply that non-surgical treatment for diabetes is possible [37].

An interesting study would be to determine the participation of pancreatic delta-cells in the maintenance of beta-cell mass after RYGB surgery since a recent study demonstrated that delta-cells become insulin-expressing cells after the ablation of insulin-secreting betacells in human islets [38] (Figure 2). This should be investigated in the future.

Other cell types, such as pancreatic epsilon-cells, do not seem to be affected after RYGB [15]. However, high plasma ghrelin levels were detected in obese mice six weeks after RYGB, probably due to an expansion of ghrelin-producing cells in the duodenum and stomach of these mice [39].

On the contrary, the plasticity of the pancreatic alpha-cell population under stressful circumstances is well known. Pregnancy or intermittent fasting are capable of enhancing the alpha-cell mass in mice [40,41]. Some factors related to the functionality of hepatic glucagon receptors (GCgr) have been proposed as brakes and regulators of alpha-cell population expansion in animal models [42]. In this sense, RYGB is also able to cause an increase in the alpha-cell population in mice six months after the operation, including a loss of beta identity markers such as PDX-1 and a gain of alpha-cell markers such as ARX in the islets (Figure 2). All of this suggests long-term trans-differentiation of beta-cells into alpha-cells after surgery [25].

This brings us to long-term relapse of diabetes again. Like SG, the outcomes of RYGB published in relevant trials have shown a progressive worsening of diabetes-related parameters such as glycated hemoglobin, reaching a 50% relapse in diabetes at five years [2]. Patel et al. proposed weak beta-cell function and peripheral insulin resistance as possible causes of relapse after RYGB [43]. An decrease in beta-cell mass and an increase in alphacell mass could explain this, but what is the mechanism that triggers trans-differentiation? Hyperinsulinism and subsequent hypoglycemia have been a problem after RYGB but also may be the answer [44]. In this sense, RYGB seems to cause an extreme requirement and stressful situation to the beta-cell population, triggering conversion to alpha-cells [45]. According to this, a study in patients reported hyperinsulinism but elevated postprandial glucagon secretion after RYGB. However, the same study did not report extremely increased beta cell function [46]. The landscape is complex and exciting and could be a good line of research to improve the efficiency of these surgeries in the remission of diabetes.

#### **4. Conclusions**

SG and RYGB are a therapeutic option not only for overweight but also for diabetes. The effects of these surgeries on enterohormonal levels have been extensively studied but on another level, further research on endocrine pancreatic cell populations is also needed. Nevertheless, it seems that different pathophysiological mechanisms underlie each of these surgeries, at least in reference to their pancreatic involvement. This is a complicated issue in humans. However, a better understanding of the mechanisms and cellular dynamics governing these populations after these two surgeries would allow us to limit hypoglycemic episodes, the relapse of diabetes over time or even the development of pharmacological alternatives to the use of bariatric/metabolic surgery.

**Author Contributions:** Writing—original draft preparation, J.F.-V., A.C.-R., A.D.G., J.B.d.l.R., A.R.-G., D.A.-G. and C.C.-M.; writing—review and editing, G.-M.P.-A. and J.-A.P.-O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The scientific articles consulted for the preparation of this review were obtained from the following databases: Pubmed, Science Direct, Google Academic and Rodin: (UCA Institutional Repository): It is a database of teaching and research objects of the University of Cadiz.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Cui Yang 1,\*, Frederik Johannes Hammer 1,2, Christoph Reissfelder 1, Mirko Otto <sup>1</sup> and Georgi Vassilev <sup>1</sup>**


**\*** Correspondence: cui.yang@umm.de; Tel.: +49-621-383-2226

**Abstract:** Obese patients are at risk of dental erosion due to micronutrient deficiency, consumption of soft drinks, gastric reflux disease and vomiting. The present study evaluates the presence of dental erosion in obese patients before and after bariatric surgery using the BEWE (basic erosive wear examination) scoring system. A total of 62 patients with severe obesity were included in the analysis, 31 in the control group (without bariatric surgery) and 31 in the surgery group (after bariatric surgery). BEWE scores did not vary between groups. Vitamin D deficiency was detected in 19 patients in the control group and three in the surgery group (*p* < 0.001). The serum calcium and vitamin D values were significantly higher in the surgery group (*p* = 0.003, *p* < 0.001 consecutively). All patients after bariatric surgery showed compliance with supplements, including vitamin D and calcium daily. Patients after bariatric surgery were less likely to drink soft drinks regularly (*p* = 0.026). Obese patients, before or after bariatric surgery, are at risk for erosive dental wear. However, with sufficient education prior to surgery and consistent intake of vitamin and mineral supplements, significant erosive dental wear after bariatric surgery could be avoided. Regular dental examination should be included in the check-up and follow-up program.

**Keywords:** obesity; follow-up; substitution; micronutrient deficiency; dental health; RYGB; VSG; sleeve gastrectomy

#### **1. Introduction**

Among patients with poorly managed obesity, metabolic/bariatric surgery has been proven to be the most effective and durable therapy for obesity [1]. Obesity and bariatric surgery have been shown to be associated with a higher risk for dental wear [2,3], which is multifactorial: frequent consumption of soft drinks is associated with obesity and dental problems [4–6]; unhealthy food choices might have led to micronutrient deficiency before surgery, e.g., iron, vitamin D and calcium [7–10]; aversion to special foods and taste changes were often reported after surgery [11–14]; and the reduced intestinal absorptive surface area affected hormonal mediators, which can lead to micronutrient deficiency postoperatively, including lower vitamin D and calcium in serum [15–17].

Erosive tooth wear is defined as a chemical–mechanical condition with an increasing prevalence worldwide, which results in a loss of hard dental tissue [18,19]. The erosion of enamel, the outer surface layer of the teeth, leads to an exposure of the dentin [20]. Unprotected dentin creates hypersensitivity to physical stimuli, such as heat and cold [21]. With progressive degradation processes, exposure of the pulp and thus avitalization of the teeth can occur [20]. As an important biological factor, saliva buffers the pH in the oral cavity and decelerates the process of dental erosion [22]. Low pH is a risk factor for hypersensitivity and erosion: dietary acids (e.g., soft drinks) have been associated with erosive tooth wear [23,24], and a clear impact on erosion prevalence was found in patients with gastro-esophageal reflux (GERD) and eating disorders associated with vomiting [25].

**Citation:** Yang, C.; Hammer, F.J.; Reissfelder, C.; Otto, M.; Vassilev, G. Dental Erosion in Obese Patients before and after Bariatric Surgery: A Cross-Sectional Study. *J. Clin. Med.* **2021**, *10*, 4902. https://doi.org/ 10.3390/jcm10214902

Academic Editor: David Benaiges Boix

Received: 24 September 2021 Accepted: 21 October 2021 Published: 24 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

In contrast, normal calcium concentration is considered as a major protective factor in determining the erosive potential [19].

GERD and vomiting are potential obesity-related problems and complications after bariatric surgery [26]. Considering the fact that patients are at further risk for calcium and vitamin D deficiency after bariatric surgery, it is suggested that the oral impactions of bariatric surgery might have consequences such as dental erosion. A recent review based on five Brazilian studies concluded that patients undergoing bariatric surgery had a higher incidence of dental complications [27]. The changes of saliva after bariatric surgery are also contradictory in the literature. Robust evidence, especially for dental erosion, is still lacking.

This study aimed to compare the dental erosion in obese patients before and after bariatric surgery. Potential dietary and health factors associating with erosive tooth wear were also evaluated.

#### **2. Materials and Methods**

#### *2.1. Study Design*

The study was designed as a single-center, cross-sectional analysis observing the effects of bariatric procedures on the severity of tooth erosion. The study was approved by the university faculty ethics committee and institutional review board (#2020-598N) and was conducted at a university hospital. The trial is registered in the German Clinical Trials Register (DRKS00025580).

#### *2.2. Sample Size Calculation and Inclusion*

Based on the described changes of dental wear and salivary flow after bariatric surgery in a previous study [28], we used G \* Power sample size calculator [29] and set alpha to 0.05 and power to 0.9, resulting in a sample size of 30 participants per group.

The control group was composed of patients who possessed poorly managed obesity (body mass index (BMI) ≥ 35 kg/m<sup>2</sup> and one or more comorbidities (e.g., diabetes, arterial hypertension, or sleep apnea), or BMI ≥ 40 kg/m2) who presented for bariatric surgery in the outpatient clinic; the surgery group consisted of patients who had already underwent bariatric surgery at least 3 months previously. Participation for all patients was predicated on their written informed consent.

#### *2.3. Dental Examination*

#### 2.3.1. BEWE Score

The BEWE (basic erosive wear examination) is a simple tool designed to assess the level of dental erosion [30]. The teeth are divided into sextants. Only the value of the tooth surface with the highest BEWE value per sextant is documented. A value from 0 to 3 is determined in each sextant. Table 1 shows the criteria for sextant scores from 0 to 3, which are summed to obtain a cumulative score (0–18), which is the basis for determining interventions. The BEWE categories define the severity of erosion in 4 groups: group 0: 0–2 points; group 1: 3–8 points; group 2: 9–13 points; and group 3: 14–18 points. The BEWE categories can be further divided into the low-risk group (group 0 and 1) and the high-risk group (group 2 and 3). Patients in the high-risk group require dental health interventions.

**Table 1.** BEWE (basic erosive wear examination) scores and criteria.


#### 2.3.2. Sialometry

To determine the saliva flow rate, a measurement of the unstimulated saliva production within 5 min as the "spitting method" was used [31]. Using the unstimulated

method, the naturally produced saliva is gathered in the floor of the mouth and spat into a collecting tube at certain time intervals. A hyposalivation is defined as salivary flow below 0.25 mL/min.

Blood samples were collected on the same day of dental examination. Vitamins and minerals, including vitamin D and calcium in serum, were measured in a routine diagnostic setting.

#### *2.4. Questionnaire*

Using an investigative questionnaire, the following data were collected: sociodemographic data (age, gender, education level, income, and migrant status), preexisting comorbidities (e.g., diabetes, arterial hypertension, and thyroid disease), dietary habits (soft drinks, eating frequencies, smoking, alcohol consumption and eating disorders), gastrointestinal discomforting (regurgitation, gastroesophageal reflux and vomiting) and dental health (hypersensitivity, history of dental or periodontal disease) as well as dental health awareness (dentist visit frequency and oral hygiene).

#### *2.5. Statistical Analysis*

All statistical calculations were performed with the RStudio Version 1.2.5042, "Double Marigold" (Boston, MA, USA) and Python 3.9.5. (Wilmington, DE, USA) For quantitative variables, the mean and standard deviations were assessed. For qualitative factors, absolute and relative frequencies were given. For non-normally distributed data, the Wilcoxon rank-sum test was used. Pearson correlation coefficient measured correlations between two metric variables, Spearman correlation between an ordinal and a metric variable, and Chi2-test between a nominal and an ordinal variable. In general, the result of a statistical test was considered statistically significant for a *p*-value < 0.05.

#### **3. Results**

The study was conducted at our university hospital between January and March 2021. All operations were performed laparoscopically by two of the co-authors (M.O. and G.V.) and the dental examinations were carried out by a dentist (co-author F.J.H.).

#### *3.1. Demographics*

Obese patients who presented for bariatric surgery or a follow-up after bariatric surgery were included. Sixty-two patients were enrolled (thirty-one in each group), with an average age of 40 years old. The majority of the participants are females. Participants in the control group are younger and more obese. Other sociodemographic characteristics did not differ between groups (Table 2).

**Table 2.** Sociodemographic characteristics of 62 patients.


BMI, body mass index; n.a., not available; n.s., not significant; SD, Standard Deviation. Significant *p*-values are highlighted in bold (Wilcoxon, Chi2 and Fisher's exact tests, *p* < 0.05). The education was classified based on ISCED 2011 (International Standard Classification of Education [32]). The income was classified by statistic data in Germany [33,34].

#### *3.2. Oral Health Related Parameters*

Patients after bariatric surgery were less likely to drink soft drinks regularly (*p* = 0.026) and reported less GERD (*p* = 0.012). Other parameters, including hypersensitivity of the teeth, eating disorders and vomiting, were similar among the groups (Table 3).

**Table 3.** Oral health related parameters.


GERD, gastroesophageal reflux disease; n.s., not significant. Significant *p*-values are highlighted in bold (Chi2 and Fisher's exact tests).

#### *3.3. Dental Examination and Level of Serum Vitamin D/Calcium*

All patients after bariatric surgery have taken the recommended supplements, including vitamin D and calcium citrate, daily. Among all patients, three patients had a hypocalcaemia (<2.18 mmol/L), who were all in the control group. No hypocalcaemia was found in the surgery group. Vitamin D deficiency (<20 μg/L) was detected in 19 patients in the control group and three in the surgery group (*p* < 0.001). The serum calcium and vitamin D values were significantly higher in the surgery group (*p* = 0.003, *p* < 0.001 consecutively).

BEWE scores, BEWE category and BEWE risk groups did not vary between groups. Of the preoperative obese patients, 22.6% were classified in BEWE category 2 and 3, therefore they had a high risk for further exposure of the pulp and consequential avitalization of the teeth (Figure 1). In the surgery group, the percentage was even higher (32.2%), but the difference was not significant. Neither was there a significant difference in the salivary flow (Table 4).

#### *3.4. Correlation*

Sociodemographic parameters, such as age, education level, migration status and comorbidities, did not show significant correlations with BEWE scores, risk and salivary flow.

#### *3.5. VSG versus RYGB*

In the surgery group, seven patients underwent a vertical sleeve gastrectomy (VSG), 21 underwent a Roux-en-Y gastric bypass (RYGB) and three others underwent a singleanastomosis duodeno-ileal bypass (S.A.D.I-S) or bilio-pancreatic diversion with duodenal switch (BPD-DS). The frequency of GERD, vomiting, BEWE scores, BEWE categories, risk, salivary flow, salivary flow categories, calcium and vitamin D did not differ between patients who underwent VSG and RYGB.

**Figure 1.** Portion of BEWE categories in control and surgery group.


**Table 4.** Results of dental examinations.

BEWE (basic erosive wear examination); n.s., not significant. All values are shown as means and standard deviation or frequency and percentage. Significant *p*-values (*p* < 0.05) are highlighted in bold.

#### *3.6. Short-Term Follow-Up versus Long-Term Follow-Up*

In the surgery group, the period between bariatric surgery and dental examination was 11 (3–142) months. Patients with a shorter follow-up (<11 months) had significantly higher calcium levels in the serum (*p* = 0.014). Vitamin D, BEWE scores and salivary flow did not differ between patients with a short-term (<11 months) and long-term (>11 months) follow-up.

#### **4. Discussion**

The prevalence of obesity is increasing, along with the number of bariatric surgeries. Accordingly, the side effects of obesity and bariatric surgery are gaining growing attention. Several factors associated with bariatric surgery might lead to dental health problems: micronutrient deficiency, as vitamin D and calcium deficiency accompanies the great severity of oral disease [35,36]; increased prevalence of gastroesophageal reflux and vomiting, which lowers the pH in oral cavity, and is consequently a major risk factor for

erosive dental wear [23,24]; and the postoperative, recommended, small yet frequent meals (4–6 meals/day), which shorten the regeneration period for the saliva [28,37], which is of great importance for the hard tissue protection.

The current study confirmed that a significant number of obese patients are at a high risk for erosive dental wear and can experience further exposure of the pulp and thus avitalization of the teeth. However, the condition of dental wear did not worsen significantly after bariatric surgery. This is not in line with the limited data in the literature which have evaluated the effect of bariatric surgery on dental erosion. Quintella et al. reviewed five Brazilian studies and concluded that patients undergoing bariatric surgery had a higher incidence of dental wear [27]. Of these studies, one focused on erosive damage, and showed more severe dental erosion in patients after bariatric surgery [38]. The divergence of conclusions might be multifactorial. Firstly, the entire population in our surgery group have taken recommended supplements, including calcium citrate and vitamin D3, which was confirmed by the significantly higher serum levels of calcium and vitamin D after surgery. Calcium and vitamin D are known to have a protective effect on hard tooth tissues [39]. The recommended supplement of vitamins and minerals was not mentioned in the Brazilian studies. Secondly, GERD and vomiting were not increased after bariatric surgery in our observation. On the contrary, GERD was significantly less reported in the surgery group, which can be explained by two facts: postoperative patients were less obese (significant lower BMI) and therefore possessed a decreased risk for GERD [40], and, in the majority of the postoperative patients, RYGB was performed. After RYGB, gastric reflux remission is more frequently observed than after VSG [41]. Thus, RGYB could be a protective factor for dental health. Thirdly, the daily consumption of soft drinks, another potential factor to lower pH in oral cavity, was significantly reduced in our surgery group. Previous investigations showed that soft drink consumption can contribute to detrimental oral health, especially due to the erosive potential [6]. Postoperative taste changes might explain the altered dietary habits [11,13]. Moreover, patients were required to take part in a minimally 6-month long, multimodal concept, including intensive consultation by nutritional therapist before the surgery. Nutritional therapists assess and correct their nutritional status and, more importantly, educate the patients on how to establish healthy dietary habits (fewer soft drinks, less carbohydrates, and more proteins, etc.). The presurgical education seemed to have a lasting effect in the patients, which was not mentioned in the Brazilian studies. Fourthly, the spans between the bariatric surgery and the survey differ. Interestingly, studies confirming a worsening of dental wear in patients after bariatric surgery were performed with a relatively short flow-up (3–6 months) [28,42]. Even with a longer follow-up span (with a median of 11 months), no significant difference was detected in our study population. Moreover, patients with a longer follow-up did not present with a higher risk for erosive dental wear than patients with a short-term follow-up in our study. It can be assumed that the risk of erosive dental wear would not increase if patients were compliant with supplements and follow-up.

Saliva is an important biological factor, which buffers the pH in the oral cavity and decelerates the process of dental erosion [22]. The impact of bariatric surgery on saliva production remains unclear. Some studies showed an improvement in the salivary flow rate, while others found no change or even a worsening in saliva production after the surgery [43]. In our study, the salivary rate did not differ between the groups. The inconsistency might be partially due to different measuring methods. We preferred the unstimulated spitting method to the stimulated measurement, since the unstimulated saliva flow, which occurs around 14 to 16 h per day, is primarily responsible for the maintenance of oral health and the protection of our teeth. The stimulated saliva flow rate, on the other hand, embodies the functional capacity of the salivary glands and is only present for around two hours a day [31]. Furthermore, our measurement was performed in two different groups, and interindividual differences might preexist for salivary production.

To the best of our knowledge, this is the first study evaluating erosive dental ware in obese patients before and after bariatric surgery in Europe. Bariatric surgery might be associated with risks of erosive dental wear due to multiple factors. However, with sufficient education prior to surgery, consistent intake of vitamin and mineral supplements, and regular follow-ups, significant erosive dental wear after bariatric surgery could be avoided. Regular dental examination should be included in the follow-up program after bariatric surgery. A remineralization solution might help to prevent dental erosions from occurring [44].

#### **5. Limitations**

The majority of patients in the surgery group received a Roux-en-Y gastric bypass. The results might be different in patients after sleeve gastrectomy. Overall, there are more female than male participants in our study. Patients were not perfectly paired in both groups due to different BMIs and ages. To minimize the bias, dental examination should be performed in patients indicated for bariatric surgery before and after surgery in further trials.

#### **6. Conclusions**

Obesity and bariatric surgery might be associated with risks for erosive dental wear due to multiple factors. However, with sufficient education prior to surgery and consistent intake of vitamin and mineral supplements, significant erosive dental wear after bariatric surgery could be avoided. Regular dental examination should be included in the check-up and follow-up program in obese patients before and after bariatric surgery.

**Author Contributions:** Conceptualization, C.Y., F.J.H., M.O. and G.V.; methodology, C.Y., F.J.H. and G.V.; formal analysis, C.Y.; investigation, F.J.H.; data curation, C.Y.; writing—original draft preparation, C.Y. and F.J.H.; writing—review and editing, C.R., M.O. and G.V.; supervision, C.R., M.O. and G.V.; project administration, C.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the university faculty ethics committee and institutional review board (#2020-598N).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Assim A. Alfadda 1,2,\*, Mohammed Y. Al-Naami 3, Afshan Masood 1, Ruba Elawad 1, Arthur Isnani 1, Shaik Shaffi Ahamed <sup>4</sup> and Nora A. Alfadda <sup>5</sup>**


**Abstract:** Background: Obesity is considered a global chronic disease requiring weight management through lifestyle modification, pharmacotherapy, or weight loss surgery. The dramatic increase in patients with severe obesity in Saudi Arabia is paralleled with those undergoing bariatric surgery. Although known to be beneficial in the short term, the long-term impacts of surgery within this group and the sustainability of weight loss after surgery remains unclear. Objectives: We aimed to assess the long-term weight outcomes after bariatric surgery. Setting: The study was conducted at King Khalid University Hospital (KKUH), King Saud University Medical City (KSUMC) in Riyadh, Saudi Arabia. Methods: An observational prospective cohort study on adult patients with severe obesity undergoing bariatric surgery (sleeve gastrectomy (SG) or Roux-en Y gastric bypass (RYGB)) during the period between 2009 and 2015 was conducted. Weight loss patterns were evaluated pre- and post-surgery through clinical and anthropometric assessments. Absolute weight loss was determined, and outcome variables: percent excess weight loss (%EWL), percent total weight loss (%TWL), and percent weight regain (%WR), were calculated. Statistical analysis using univariate and multivariate general linear modelling was carried out. Results: A total of 91 (46 males and 45 females) patients were included in the study, with the majority belonging to the SG group. Significant weight reductions were observed at 1 and 3 years of follow-up (*p* < 0.001) from baseline. The %EWL and %TWL were at their maximum at 3 years (72.4% and 75.8%) and were comparable between the SG and RYGB. Decrements in %EWL and %TWL and increases in %WR were seen from 3 years onwards from bariatric surgery until the study period ended. The yearly follow-up attrition rate was 20.8% at 1 year post-surgery, 26.4% at year 2, 31.8% at year 3, 47.3% at year 4, 62.6% at year 5, and 79.1% at end of study period (at year 6). Conclusion: The major challenge to the successful outcome of bariatric surgery is in maintaining weight loss in the long-term and minimizing weight regain. Factors such as the type of surgery and gender need to be considered before and after surgery, with an emphasis on the need for long-term follow-up to enssure the optimal benefits from this intervention.

**Keywords:** weight regain; bariatric surgery; obesity; long-term follow-up; weight loss

#### **1. Introduction**

Obesity has become a worldwide epidemic: according to the 2016 World Health Organization statistics, 1.9 billion adults were overweight, and more than 650 million were

**Citation:** Alfadda, A.A.; Al-Naami, M.Y.; Masood, A.; Elawad, R.; Isnani, A.; Ahamed, S.S.; Alfadda, N.A. Long-Term Weight Outcomes after Bariatric Surgery: A Single Center Saudi Arabian Cohort Experience. *J. Clin. Med.* **2021**, *10*, 4922. https:// doi.org/10.3390/jcm10214922

Academic Editor: David Benaiges Boix

Received: 12 October 2021 Accepted: 22 October 2021 Published: 25 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

obese [1]. This increase in the number of individuals with overweight and obesity is also reflected in the Saudi population [2,3], where the estimated rate of overweight and obesity is 24.1% in men and 33.5% in women [3]. Obesity management aims at achieving weight loss, utilizing a multifactorial stepwise approach consisting of behavioral therapy, lifestyle and dietary interventions, and medical pharmacotherapy. Weight loss through bariatric surgery is an adjunct to the former strategies in patients who did not benefit from them or who have additional comorbidities associated with obesity.

Presently, bariatric surgery is globally considered among the most effective management modalities for patients with obesity. The Saudi clinical practice guidelines, similar to the American Society for Metabolic and Bariatric Surgery, recommends that bariatric surgery should be conducted for patients with obesity who have a body mass index (BMI) ≥40 or ≥35 kg/m2 and the presence of comorbidities [4] to diminish the risk of the associated comorbidities and to improve quality of life [5,6]. In Saudi Arabia, the number of patients with obesity who undergo bariatric surgery has noticeably grown [7], and ≥15,000 procedures are estimated to be performed annually [8]. These procedures are known to be effective and lead to major weight loss, with the maximum weight loss occurring at 12–18 months post-surgery [9]. The Swedish Obese Subjects (SOS) study, the largest non-randomized intervention trial that compared weight loss outcome over 10 years, reported a maximal total body weight change in patients after 1 year after receiving Rouxen-Y gastric bypass (RYGB) and vertical banded gastroplasty surgeries, with reductions of 38 ± 7% and 26 ± 10%, respectively [5]. A successful outcome of bariatric surgery is one that achieves a loss of 50–70% of excess weight (EWL) or the 20–30% loss of the patient's initial weight, or a BMI < 35 kg/m<sup>2</sup> [9,10]. Although many studies have documented the competence and effectiveness of bariatric surgery in reducing excess weight in the short term, mid- and long-term studies have reported weight regain as an index for failure of the surgery, regardless of the surgical procedure [5,11]. According to the SOS study, patients were noted to have regained around 20–25% of their lost weight at 10 years post-surgery. A weight regain of 12% total body weight was observed in patients who underwent RYGB, while those reported for sleeve gastrectomy (SG) were variable, ranging from 6% at as early as two years post-surgery to 76% at six years post-surgery [12]. Previous studies have addressed different factors that contribute to weight regain and the failure of bariatric surgery, including dilation of the gastric pouch or gastro-jejunal anastomosis, pre-surgery BMI, eating behaviors, increases in energy intake, level of physical activity, the patient's lack of commitment to follow-up visits, and psychological factors [13–18].

Obesity, similar to other chronic diseases, persists for prolonged durations and requires a continuous close follow-up to re-assess the efficacy of treatments, including bariatric surgeries. Compared to the number of surgeries being performed, there are few mid-longterm studies assessing the effectiveness and changes in weight loss. This is even more true in case of the Saudi population, where only a few studies have addressed bariatric surgery outcomes [4,7,8,19,20] and fewer have evaluated the long-term weight loss outcomes. In this study, we aim to observe and evaluate the weight evolution pattern in Saudi patients during a six-year follow-up period following bariatric surgery.

#### **2. Materials and Methods**

#### *2.1. Study Design, Setting and Subjects*

An observational prospective cohort study of adult patients with obesity (age ≥ 18 years) who underwent SG and RYGB was conducted at King Khalid University Hospital (KKUH), King Saud University Medical City (KSUMC) in Riyadh, Saudi Arabia, between 2009 and 2015. All procedures and protocols were reviewed and approved by the Research Ethics Committee of the College of Medicine, King Saud University. Written informed consent was obtained from all participants. The bariatric procedures were conducted under the supervision of a single bariatric surgeon. Patients who underwent bariatric surgery and who had follow-up visits at the Obesity Clinic at KKUH and the Obesity Research Center

for a period of 6 years were included in the study. Their pre and postoperative clinical data and anthropometric measurements were collected and recorded during follow-up.

Weight (in kilograms) was measured in light clothing and without shoes to the nearest 0.1 kg. Height was measured using a stadiometer, and BMI was calculated. The variables analyzed were age, sex, BMI, absolute weight loss, %EWL, percent total weight loss (%TWL), and percent weight regain (%WR).

#### *2.2. Calculated Variables*

We calculated the absolute weight loss as ((follow-up weight − pre-surgery weight/presurgery weight) × 100) for all of the different time points. The outcome variables in the study included %EWL, %TWL, and %WR. The %EWL was calculated as ((pre-surgery weight − follow-up weight)/(operative excess weight)) × 100, where the operative excess weight equaled (pre-surgery weight − ideal weight) and where the ideal weight was based on the metropolitan tables [21]. A %EWL of ≥50% represented successful weight loss; a %EWL of ≤50% was considered as a failure [22]. In other studies, the rates indicating failure were reported to be ≤25% at ≥5 years [23,24]. Given the variability of %EWL depending on the definition of the ideal body weight, we used %TWL, as it is reported to be less influenced by BMI and other anthropometric measures. The %TWL was calculated as follows: ((preceding year weight − current weight)/preceding year weight) × 100 [25]. The %WR, which is the percentage of weight regained from the nadir weight (lowest measured post-surgery weight), was calculated using the following formula: ((current weight − nadir weight))/(pre-surgery weight − nadir weight) × 100), where ≥25% weight gain from nadir was considered to be excessive weight regain [22]. After surgery, the patients were prospectively followed up with in the clinic for 6 years. A comprehensive anthropometric measurement (weight, height, and BMI) was performed before surgery and post-surgery at each time point annually until the end of the study period. The yearly attrition rate (in %) was derived by dividing the number of withdrawn participants (calculated as the number of participants retained in the study subtracted from the total number of participants originally included in the study at pre-surgery) by the number of participants originally included in the study × 100 [26].

#### *2.3. Statistical Analysis*

Data were analyzed using the SPSS 24.0 Advanced statistics module (IBM Inc., Chicago, IL, USA). Categorical variables (gender and type of surgery) were reported as numbers and percentages, whereas continuous variables (age, anthropometric measurements, %EWL, %TWL and %WR) were reported as mean, standard deviation, and range. The mean weights for all patients from pre-surgery and across the six follow-up time points were graphically presented according to gender and type of surgery.

Changes in the mean values of three outcome variables: %EWL, %TWL, and %WR, over the six time points were compared using the repeated measures analysis of variance (ANOVA), and F values were reported to represent the systematic variance of %EWL, %TWL, and %WR across the six time points. Repeated measures of analysis for the outcome variables (%EWL, %TWL, and %WR) with each of the independent variables (gender and type of surgery) were also conducted.

A generalized linear mixed model for repeated measures for univariate and multivariate analysis were used to evaluate the changes in the quantitative outcome variables, which were observed at the six observation time points (1 to 6 years post-surgery). The fixed effects used in the model were time points, gender, and type of surgery. Akaike-corrected information criteria were used to identify the model of best fit. The least significant difference criterion was used to calculate the adjusted *p*-values in a pairwise comparison of the mean values. A *p*-value of <0.05 was used to report the statistically significant results.

#### **3. Results**

A total of 91 (50.5% male) patients with the mean age of 33.3 ± 9.7 years who underwent bariatric surgery were included in the study at baseline. The mean pre-surgery weight and BMI were 134.4 ± 33.8 kg and 49.7 ± 9.9 kg/m2, respectively. Table <sup>1</sup> shows a detailed demographic profile of the total study population.

**Table 1.** Overall demographic data and baseline characteristics of patients with obesity who underwent bariatric surgery. The values are represented as mean ± SD (standard deviation).


The overall annual follow-up rate of the patients was 79.1% (33 males/39 females) at 1 year, 81.3% (35 males/39 females) at 2 years, 68.1% (30 males/33 females) at 3 years, 54.9% (27 males and 23 females), 42.9% (17 males and 17 females), and 20.9% (7 males, 12 females) at 6 years. The yearly follow-up attrition rate was 20.8% at year 1 post-surgery, 26.4% at year 2, 31.8% at year 3, 47.3% at year 4, 62.6% at year 5, and 79.1% at end of the study period, i.e., year 6 post-surgery.

To characterize the weight change patterns in our cohort, a subgrouping of the patients according to the type of surgery and gender was conducted. Based on the type of surgery, the patients in the RYGB group accounted for 31.9% of patients (*n* = 29, 11 males and 18 females) while the SG group comprised 68.1% of patients (*n* = 62, 35 males and 27 females). The maximum mean weight-loss percentage in the RYGB and SG groups was seen at 3 years post-surgery and was similar (54.3% and 54.4%, respectively). Increments in weight were observed in both the bariatric surgery groups beyond the 3-year follow-up period. Both the groups also showed an increasing weight gain trend from year 4 post-surgery and onwards (Figure 1).

Based on gender, we noticed that the maximum mean percentage weight loss occurred at 3 years post-surgery by as much as −65.07% in males, while it was −43.48% at 4 years post-surgery for females (Figure 2). The rate of weight regain was seen to increase gradually from 3 years post-surgery onwards until the end of the study period. Significant weight regain (defined as ≥25% weight gain from nadir weight) was seen in 53.3% of the patients after 6 years.

**Figure 1.** Graphical representations of changes in weight (%), with the error bars representing the ±2 SD during the 6-year follow-up period for all patients who underwent bariatric surgery (blue line), for those who underwent SG (red line), and those who underwent RYGB (green line).


**Figure 2.** Graphical representation of weight change (%), with the error bars representing the ±2 SD, according to gender of the patients post bariatric surgery during the 6-year follow-up period. The number of patients at each time point and the mean percentage weight loss in male and female participants is shown. SD-standard deviation.

Three outcome variables (%EWL, %TWL, and %WR) were considered for the analysis. A total of 314, 308, and 98 repeated measures were used in the analysis for these outcome variables. Table 2 shows the comparison of the mean values of %EWL, %TWL, and %WR across the six observation time points, and Table 3 shows the comparison of mean values of %EWL, %TWL, and % WR across the observation time points in male and female patients and in those who had undergone RYGB and SG surgeries. A similar comparison is shown for %EWL (Figure 3A–C), %TWL (Figure 3D–F), and %WR (Figure 3G–I).


**Table 2.** Comparison of repeated measure (%EWL, %TWL, and %WR) mean (SD) values of study subjects across the six time points and the difference between each time point and at the end observation time point.

%EWL—percentage excess weight loss, %TWL—percentage total weight loss, %WR—percentage weight regain.

**Table 3.** Comparison of repeated measure (%EWL, %TWL, and %WR) mean values across the six time points in male and female subjects and in relation to type of surgery.


*3.1. Univariate Analysis: Generalized Linear Mixed Effects Model Analysis for the Outcome Variables %EWL, %TWL and %WR for Each of the Independent Variables (Time Points, Type of Surgery and Gender)*

The univariate repeated measures and generalized linear mixed effects model for the outcome variables %EWL, %TWL, and % WR across the six time points showed statistically significant differences (F = 10.82, *p* < 0.0001; F = 43.99, *p* < 0.0001; F = 2.72; *p* = 0.034) (Table 2).

The mean %EWL values were significantly increased at 3 and 4 years post-surgery when compared to the mean values at 6 years post-surgery (*p* < 0.0001), where the coefficients at 3 (19.04, *t* = 3.21, *p* = 0.001) and 4 (21.02, *t* = 2.74, *p* = 0.006) years indicate that %EWL increased by 19.04 and 21.02 units when compared to the mean value of %EWL at 6 years post-surgery.

**Figure 3.** The figure shows a line graph with 95% confidence intervals and depicts the changes in the estimated mean of outcome variables with the predictors over the study duration at the different time points. The changes in %EWL with (**A**) different time points, (**B**) time points and type of surgery, and (**C**) time points and gender; the changes in %TWL with (**D**) different time points, (**E**) time points and type of surgery, and (**F**) time points and gender; and the changes in %WR with (**G**) different time points, (**H**) time points and type of surgery, and (**I**) time points and gender are shown.

The mean %TWL values were significantly higher at 1 and 2 years post-surgery when compared to the mean values at 6 years post-surgery (*p* < 0.0001), where the coefficients at 1 (23.35, *t* = 7.90, *p* < 0.0001) and 2 years (16.57, *t* = 5.08, *p* < 0.001) showed that the %TWL mean values increased by 23.35 and 16.57 units when compared to the mean value of %TWL at 6 years post-surgery.

The mean %WR values were significantly decreased at 2 (−20.18, *t* = −2.67, *p* = 0.009) and 3 (−17.06, *t* = −2.85, *p* = 0.005) years post-surgery when compared to the mean values at 6 years post-surgery, indicating that the %WR mean values decreased by 20.18 and 17.06 units when compared to the mean value of %WR at 6 years post-surgery.

The comparison of the mean values of %EWL in each of the surgery groups (RYGB and SG) across the six time points showed highly statistically significant differences (F = 2.89, *p* = 0.018 and F = 9.27, *p* < 0.0001), respectively (Table 3). The mean values of %EWL in patients who underwent RYGB surgery at 3 years was increased by 24.82 units (*t* = 2.24, *p* = 0.027), while it increased by 15.89 units (*t* = 2.14, *p* = 0.033) in those who underwent SG when compared to the mean value of %EWL at 6 years post-surgery.

Similarly, high statistically significant difference was observed for the mean values of % TWL in each of the surgery groups (RYGB and SG) across the six time points (F = 20.53; *p* < 0.0001 and F = 27.63; *p* < 0.0001) (Table 3). The mean values of %TWL in patients who underwent RYGB surgery was higher at 1 and 2 years by 24.64 units (*t* = 4.85, *p* < 0.001) and 17.47 units (*t* = 3.19, *p* = 0.002), while it was higher by 22.87 units (*t* = 6.90, *p* < 0.001) and by 16.15 units (*t* = 4.24, *p* < 0.001) in those who underwent SG when compared to the mean value of %TWL at 6 years post-surgery. Hence, it can be inferred that the mean %TWL

values were significantly lower after 6 years in patients who had undergone either surgery when compared to the mean values at 1 and 2 years post-surgery.

However, for the mean values of % WR, a non-significance was observed in each of the surgery groups (RYGB and SG) across the six time points (F = 2.28; *p* = 0.080 and F = 1.27, *p* = 0.294) (Table 3). However, the pairwise compassion of the time points shows that patients who underwent RYGB surgery had significantly decreased mean values of %WR at 2 and 3 years by −20.82 units (*t* = −2.64, *p* = 0.012) and 16.83 units (*t* = −2.24, *p* = 0.031), while those who underwent SG showed a significant decrease in the mean value of %WR at 3 years by 19.55 units (*t* = −2.10, *p* = 0.041) when compared to the mean value of %WR at 6 years post-surgery.

The comparison of the mean values of %EWL in each of the gender groups (male and female) across the six time points showed highly statistically significant differences (F = 7.22; *p* < 0.0001 and F = 4.74; *p* < 0.0001) (Table 3). The mean %EWL values in males were significantly increased at 2 (24.70, *t* = 2.45, *p* = 0.015), 3 (38.90, *t* = 10.18, *p* = 0.003), and 4 (26.61, *t* = 2.21, *p* = 0.029) years when compared to the mean value at 6 years post-surgery. In female subjects, the mean values of %EWL across the six time points were statistically significantly different, whereas the comparison of each time point with the mean values at 6 years post-surgery did not show any statistically significant differences.

Additionally, highly statistically significant differences were observed for the mean values of % TWL in both male and female subjects across the six time points (F = 16.67; *p* < 0.0001 and F = 39.15; *p* < 0.0001) (Table 3). The pairwise comparison showed that the mean %TWL values in males was significantly higher at 1 (20.69, *t* = 5.75, *p* < 0.001) and at 2 (22.10, *t* = 5.23, *p* < 0.001) years when compared to the mean value at 6 years post-surgery. This was also noted in the female subjects, and the mean values of %TWL were found to be significantly higher at 1 (25.37, *t* = 4.38, *p* < 0.001) and 2 (11.53, *t* = 4.53, *p* = 0.012) years when compared to the mean values of %TWL at 6 years post-surgery.

In each of the gender groups (male and female) the comparison of the mean values of %WR across the five time points of observations showed no statistically significant differences (F = 1.54, *p* = 0.209 and F = 1.74, *p* = 0.156) (Table 3). However, the pairwise comparison of the time points indicated that the mean %WR values in males were significantly decreased at 2 years (−20.91, *t* = −2.34, *p* = 0.024) while in females, they were significantly decreased at 3 years (−18.39, *t* = 4–2.14, *p* = 0.037) when compared to the mean values of %WR at 6 years post-surgery (Figure S1I).

#### *3.2. Multivariate Analysis: General Linear Mixed Effects Modelling for Each of the Outcome Variables, %EWL, %TWL and %WR with Independent Variables (Time Points, Type of Surgery and Gender)*

The multivariate repeated-measures generalized linear mixed effects model for the outcome variables %EWL and %TWL was significant for the overall model (F = 4.49, dff = 17, *p* < 0.0001 and F = 14.78, dff = 17, *p* < 0.0001). Significance was also noted with time points (F = 10.70, *p* < 0.001 and F = 40.05, *p* < 0.0001), gender (F = 3.53, *p* = 0.068 and F = 7.35, *p* = 0.007) and type of surgery (F = 4.14, *p* = 0.043 and F = 1.160, *p* = 0.282), time points × gender (F = 2.05, *p* = 0.071 and F = 3.36, *p* = 0.006), and time points × type of surgery (F = 0.310, *p* = 0.91 and F = 0.695, *p* = 0.628) were used as predictors of the model.

Taking the time point predictors into consideration, significant differences (*p* < 0.001) in %EWL were observed at 2 (15.06), 3 (22.12), and 4 (23.09) years and for %TWL at 1 (23.34) and 2 (17.30) years in comparison to the mean value at 6 years post-surgery. The %EWL mean values in males were observed to be statistically significant at 2 (27.97), 3 (34.15), and 4 years (29.20) when compared to the mean value at 6 years post-surgery. Males showed significantly higher %EWL at 2 years post-surgery by 25.81 units (*p* = 0.045) compared to the females, while no significant differences were noted for the coefficients for the other terms in the model. The %TWL mean values in males were observed to be significant at 1 (21.07) and 2 years (22.70) and for females at 1 (25.60) and 2 (11.85) years when compared to the mean value at 6 years post-surgery. Similar to %EWL, %TWL in males was significantly higher at 2 years post-surgery compared to females.

Considering surgery as a predictor, %EWL was significantly higher in male subjects who had undergone RYGB at 2 (21.91), 3 (29.18) and 4 (28.59) years post-surgery, while no statistically significant differences were observed in female subjects and in subjects who had undergone SG. Mean values of %TWL were statistically significant at 1 and 2 years post-surgery in patients who had undergone RYGB (23.75 and 18.73 units) and in SG the group (22.93 and 15.87) when compared to the mean values of %TWL at 6 years post-surgery.

The multivariate analysis for the outcome variable %WR with time points, gender, and type of surgery as predictors showed no statistical significance for the overall model (F = 1.10, dff = 14, *p* = 0.370). Among all of the coefficients, at 3 years post-surgery, the %WR mean value decreased by 23.59 units (*p* = 0.044) when compared to the mean value of %WR at 6 years post-surgery, whereas the coefficients for other terms in the model were not statistically significant. For the time points predictor, significant differences were observed at 2 (−20.53) and 3 years (−17.56) post-surgery when compared to the mean value at 6 years post-surgery.

No statistically significant differences in the %WR pattern were observed in male subjects in the comparison of the mean values of %WR between the pair of time points (2 to 6 years post-surgery), whereas in female subjects, the mean values of %WR at 2 (−22.02) and at 3 years (−21.08) were significantly decreased when compared to the mean value at 6 years. Statistically significant differences were observed at 3 years in patients who had undergone SG surgery, where the mean values of %WR decreased by −20.06 units when compared to the mean values at 6 years post-surgery, while no significant differences were observed in patients who had undergone RYGB surgery (Figure S1).

#### **4. Discussion**

Being a chronic disease, managing or treating severe obesity is challenging and requires constant close clinical and nutritional monitoring to reduce the recurrence of weight gain after weight loss. A large body of literature has shown that weight loss, even at a modest level of 5–10%, is beneficial and helps in the resolution of obesity associated comorbidities including, T2DM, hypertension, and fatty liver disease, among others, and in improving the overall quality of life [27–29]. Attaining a weight loss change of ≥5 and/or ≥10% of initial body weight has also been shown to be strongly correlated with a reduction in the risk of cardiovascular disease [30]. Surgical intervention techniques such as bariatric surgery were instated to provide a more permanent weight loss solution. Regardless of the type of surgery or the mechanisms by which weight loss is achieved, bariatric surgery has proven to lead to a significant amount of weight loss immediately post-surgery [31], but its long-term efficacy needs to be evaluated. The general metrics to assess the success of the surgeries includes calculating %EWL (>50%), %TWL, and %WR post-surgery [32–36]. Different studies have shown a large amount of variability within these values, which have been attributed to either the type of surgery, the preoperative BMI, and to the race and ethnicity that the patients belong to [29,30]. To date, only a limited number of studies have looked at differences in weight loss patterns across different populations and specifically in the Saudi population, where bariatric procedures are presently being performed routinely. In our present study, we conducted a longitudinal follow-up of patients undergoing bariatric surgery and observed the weight evolution patterns through annual follow ups for a period of 6 years post-surgery.

#### *4.1. Weight Evolution in the Overall Bariatric Group*

All of the patients demonstrated a uniform annual incremental increase in the weight loss outcome measures (i.e., absolute weight change, %EWL, and %TWL) as early as 1 year post-surgery, and significant changes in these parameters were seen from baseline to up to 3 years post-surgery. Our findings are different from those of the previous studies that have reported maximum weight loss to occur at 1–2 years post-surgery. The %EWL and %TWL showed a similar pattern across the cohort, where incremental increases in weight loss were seen from the baseline, peaking at the 3-year post surgery. The calculated %EWL values observed in our study population were higher than those reported by Bohdjalian et al. at the 3-year annual follow up, and while long-term weight outcomes were similar to theirs [37], they were higher than those observed in other cohorts [37,38]. Weight change patterns beyond 3 years showed a decrease in the propensity for further weight loss followed by a plateauing phase and then a slow and gradual increase in the patients' weight. Our findings are in line with the findings of the SM-BOSS study that also showed an increased BMI at 5 years [39].

Weight regain remains a major challenge in relation to the long-term success of bariatric surgery [40]. Numerous studies have previously shown a higher tendency for patients to regain their weight after an initial impressive weight loss until the midterm (>3 years), which was not substantiated in the long term (>5 years). Although weight regain is a consistent finding among studies, there are considerable inter-individual variations in the magnitude and rate of weight regain depending on factors ranging from behavioral, dietary, lifestyle, psychological, ethnic, and racial differences [22,40–43]. One of the reasons for the weight regain has been attributed to the influences of gastrointestinal hormones, including glucagon-like peptide 1 (GLP-1) ghrelin, glucose-dependent insulinotropic polypeptide (GIP), and the adipokine leptin. These hormones have been shown to regulate feelings of satiety, influence hunger and energy balance by regulating the intake and storage, and energy expenditure through the actions of the entero–hypothalamic axis [44,45]. In our study, we observed a gradual and consistent increase in the number of patients who experienced weight regain across the follow-up period. Significant weight regain (defined as ≥25% weight gain from nadir weight) was seen in 53.3% of the patients at 6 years post-surgery. The %WR was significantly higher in the 6th year in comparison to the 2nd and 3rd years of follow-up post-surgery. Our findings indicate that, similar to lifestyle and medical management of obesity, bariatric surgery is also successful in yielding short term weight loss. The notion that bariatric surgery provides a permanent solution for resolution in the long term has to be made with caution.

#### *4.2. Weight Loss Patterns between SG and RYGB*

The two most common weight loss surgical procedures that are performed are SG and RYGB, which differ in terms of the irreversible anatomical alterations created surgically at specific sites in the gastrointestinal tract (GIT). These anatomical changes in the GIT lead to many physiological and biochemical variations that produce differences in the regulation of the food and appetite, gut hormones, bile acids, and gut microbiota and consequently lead to weight loss by reducing appetite [46–49]. Few studies have claimed better weight loss outcomes with RYGB, while another study that conducted a head to head comparison between these two techniques suggests that in the long-term, only a subtle weight loss difference exists in favor of RYGB [49]. In our study, we found that the trajectories of weight loss in both the RYGB and SG demonstrated a similar trend when measured in terms of changes in weight % and %EWL from baseline up to 3 years post-surgery, at which point, the weight loss started decreasing. On the other hand, %TWL showed a steeper decline in weight loss in the RYGB group up until 4 years post-surgery compared to the SG group. Beyond the 4-year mark, both surgeries showed a similar weight pattern. Compared to SG, RYGB is considered to be the intervention that results in far greater weight loss. Our findings indicate that this assumption holds true with regard to weight loos in the short term, but the effects of both surgical procedures are similar in the long term. Our results are in line with the SLEEVEPASS and the SM-BOSS studies, which also reported no significant differences between the two bariatric methods with regard to weight loss in both the short or long term [48].

Previous studies have shown that on average, patients regain 7% of their total body weight from their lowest post-operative weight over the course of 10 years [50]. This weight regain pattern in our study was similar between the SG and RYGB groups. The RYGB group demonstrated an increase in the %WR in the 3-year post-surgery follow up, while weight regain in the SG group was seen from the 4-year follow-up onwards. Previously, the weight regain in patients who had undergone RYGB, was shown to be about 15% within two years of the surgery, which subsequently increased to 70% of patients between two and five years and to 85% at after five years post-surgery [51]. However, in our study, significant weight regain was noted at 3-year post-surgery follow up. The causes for weight regain have been shown to be due to homeostatic changes in the body post-surgery that lead to biochemical, physiological, hormonal, and metabolic adaptations to weight loss that support weight regain. These changes include perturbations in the levels of circulating appetite-related hormones and energy homoeostasis. In addition, the alterations in nutrient metabolism and subjective appetite are rather dependent on factors that depend on the physiology of the body and the metabolism. RYGB and SG induce similar changes in leptin, PYY, and GLP-1 levels, but not in the levels of ghrelin, whose levels are reduced after SG, while they are known to change over time after RYGB and may be the reason for the differences in weight regain between these methods [52]. The high prevalence of weight regain after bariatric surgery has also been a reason for an increase in the number of revisional bariatric surgeries that pose an increased surgical risk to the patient [53].

#### *4.3. Weight Loss Changes According to Gender*

In our study we also looked at the pattern of change in the weight loss patterns among the male and female participants at the different post-surgery time points. We found that males lost significantly more weight in terms of the mean values of %EWL in the 2, 3, and 4-year post-surgery follow ups, while the weight loss in the female subjects at each time point did not show any statistically significant differences with mean value of %EWL 6 years post-surgery. On the other hand, the mean %TWL values in both males and females were significantly higher at 1 and at 2 years post-surgery when compared to the mean values of %TWL at 6 years post-surgery. The differences in weight loss patterns among the genders were also seen in previous report by Tymitz et al., who showed that men had a higher absolute weight loss [54] while females showed a greater BMI loss post bariatric surgery [55]. This difference between the genders has been suggested to be the result of higher loss percentage of fat mass and an increase in fat free mass in men [56]. With regard to weight regain, we observed that males started to show an increase in weight earlier than their female counterparts, i.e., at the 3 years post-surgery. On the other hand, the weight regain seen in the females started at 4 years post-surgery, but both groups showed the same pattern at 6 years post-surgery. In their study, Meguid et al. attributed higher weight regain in females to differences in eating habits, diet, and probably a failure in developing and sustaining a large amount of plasma peptide YY levels, a hormone that regulates satiety and suppresses hunger [57]. The differences in the weight regain pattern of the males compared to the females within the two groups highlights the fact that gender is an important factor that affects the outcome of these surgeries. Additional studies to study the changes in the levels of these hormones in relation to differences in gender as well as with the type of surgery will be conducted in the future. Emphasis should be placed on regular follow ups post-surgery along with the implementation of a multidisciplinary approach to track weight regain to provide the best outcomes.

To our knowledge, the present study is the first to have looked at the long-term weight changes that occur post bariatric surgery in patients with obesity from Saudi Arabia. As previously mentioned, a lack of standardization in the reporting measures used for weight loss outcomes has been noted in the literature, which has hindered direct comparisons of weight loss among various studies. Therefore, we used two weight loss measures or outcome variables, EWL% and TWL%, in the present study to provide a broader representation of the measures of weight loss that will allow our results to be more easily compared to those of other studies [32,33,58]. Our study has a number of limitations, specifically with regard to the attrition rate and number of patients that followed up, even after being given clinical appointments. One of the reasons for this could be the increased weight regain due to unhealthy dietary habits and behavioral lifestyle practices [17], which

could have deterred the patients from attending their follow-up appointments. Although our follow-up rate is low, it is similar to that found by other groups attempting longterm follow-up studies [33,59,60]. All of the patients undergoing bariatric surgery were included without any prior stratification of their preoperative weight that could have affected the results.

#### **5. Conclusions**

The weight loss results that occur after bariatric surgery can be considered profound, as seen at 3 years post-surgery, but they are not consistent in the long run. Weight regain remains a major challenge post bariatric surgery, and long-term follow-up is required to ensure gaining the optimal benefits from this intervention.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/jcm10214922/s1, Figure S1: The figure shows a bar graph with 95% confidence intervals depicting changes in the outcome variables with the predictors over the study duration at the different time points. The changes in %EWL with (A) different time points, (B) time points and type of surgery, and (C) time points and gender are shown. The changes in %TWL with (D) different time points, (E) time points and type of surgery, and (F) time points and gender are shown. The changes in %WR with (G) different time points, (H) time points and type of surgery, and (I) time points and gender are shown. The horizontal line represents the estimated mean of the outcome variables at time point = 6. The vertical bars are the simple contrasts of the outcome variables at each level of time points minus the outcome variable at time point = 6. Significant contrasts are shaded in gold. The least significant difference adjusted significance level is 0.05.

**Author Contributions:** Conceptualization and clinical and surgical follow up, A.A.A. and M.Y.A.-N.; methodology and software, A.M. and A.I.; statistical analysis, S.S.A.; formal analysis, A.M., A.I., A.A.A. and M.Y.A.-N.; data curation, R.E., A.I. and N.A.A.; writing—original draft preparation, A.M., A.A.A. and A.I.; review and editing, A.A.A., N.A.A. and S.S.A.; funding acquisition, A.A.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was funded by the National Plan for Science, Technology and Innovation MAARIFAH), King Abdulaziz City for Science and Technology, Kingdom of Saudi Arabia. (Project No. 08-MED513-02).

**Institutional Review Board Statement:** Ethical clearance was obtained from the Research Ethics Committee of College of Medicine, King Saud University.

**Informed Consent Statement:** Written informed consent was obtained from all participants. This study was conducted at the Obesity Research Center, College of Medicine and King Khalid University Hospital, King Saud University, Riyadh, Saudi Arabia.

**Data Availability Statement:** All data generated or analyzed in the current study are included in this article.

**Acknowledgments:** We would like to thank Hadeel M. Awwad for assisting in reviewing the weight loss data and the manuscript. This work was funded by the National Plan for Science, Technology and Innovation (MAARIFAH), King Abdulaziz City for Science and Technology, Kingdom of Saudi Arabia (Project No. 08-MED513-02).

**Conflicts of Interest:** The authors have no commercial associations that might be conflicts of interest in relation to this article.

#### **References**

