Bariatric Surgeries, from Weight Loss to Weight Regain: A Retrospective Five-Years Cohort Study

Bariatric surgery is recognized as the most effective long-term treatment of morbid obesity [1‒3]. Several factors such as age, gender, genetic, obesity phenotype, lifestyle factors, as well as the type of bariatric surgery influence the percentage weight loss after surgery [4]. Laparoscopic sleeve gastrectomy (LSG), Roux-en-Y gastric bypass (RYGB), and more recently one-anastomosis gastric bypass (OAGB) have been the most popular and bariatric operations in recent years, resulting in different weight loss at different times post-surgery [5].

Recent studies have reported that bariatric surgery may lead to unsatisfactory weight loss, weight loss plateau, and even weight regain after surgery which could occur at different times after various types of surgery in each person [6, 7]. In spite of numerous studies exploring the trend and amount of bariatric surgery-induced weight loss, the mid-, and/or long-term outcome of surgery remains one of the most important concerns. Understanding the key times which is most probable to hit in weight loss plateau or weight regain in various types of surgery could be a contribution to preventive planning. Further, determining the amount and time of maximum possible weight loss, particularly in the longer term helps surgeons consider the most suitable type of surgery based on patients’ weight goal and make reasonable expectations about the surgery for patients. Thus, the main aim of the present study was to determine the key time points of attaining maximum weight loss, weight loss plateau, and weight regain after three types of bariatric surgery as well as to determine the mid-term weight loss outcome 5 years after various types of bariatric surgery.

This is a retrospective cohort study conducted in the obesity clinic of Minimally Invasive Surgery Research Center. The participants were patients with morbid obesity (BMI ≥40 and/or BMI ≥35 with more than one obesity-related comorbidities had undergone various types of bariatric procedures. The data used in this study have been obtained from the National Obesity Surgery Registry Database from March 2016 to May 2022. We included cases that had undergone one of the three types of bariatric procedures consisting of OAGB, RYGB, or LSG and had been followed up to 5 years post-surgery. The patients with conditions including post-surgery pregnancy as well as the patients with secondary obesity caused by endocrine disorders and patients with major psychiatric disorders were excluded from the analysis. The present project was approved by the Ethics Committee of Iran University of Medical Sciences with the ID number of IR.IUMS.REC 1395.95-04-140-29239.

One surgeon had done all operations. Procedure selection by the surgeon was based on patient’s age, sex, body weight, BMI, obesity comorbidities specially T2DM, endoscopic evaluation results, dietary patterns/habits (volume eating or grazing, etc.), efficacy of each procedure in weight loss, and even functional status, social supports of patients.

The weight loss outcome of bariatric surgery was evaluated as excessive weight loss (%EWL) and total weight loss (%TWL) which has been defined according to the following formula. The ideal body weight was calculated based on height and BMI = 25.

%EWL = (initial weight − postoperative weight)/(initial weight − ideal weight) × 100

%TWL = (initial weight − postoperative weight)/(initial weight) × 100

The weight regain was determined as a reduction of mean %EWL from the maximum %EWL [8]. For measuring the weight (by SECA 711 scale, Medical Measuring Systems, and Scales Factory, USA), the participants were asked to remove the heavy clothes and shoes. Their height was read while the back of the head, shoulder blades, buttocks, and heels touched the stadiometer.

Some comorbidities including impaired glucose tolerance (IGT.ei 100< FBS <125), T2DM (ei FBS ≥126), dyslipidemia, and hypothyroidism were assessed using the pre-surgery laboratory tests. Further, the history of hypertension, sleep apnea, cardiovascular disease, as well as low-back and knee pain were diagnosed and recorded by physicians. Some variables including lifestyle and socioeconomic factors, dietary pattern, smoking, alcohol or substance abuse, psychiatric or emotional status, and physical activity, which were unavailable for evaluation in this study, may play as potential confounder variable and could affect the weight loss outcomes.

To obtain a more robust analysis, we include data from all 2,567 patients available from the database. The quantitative variables were expressed as mean±SD where qualitative factors were presented as number and percentage. Statistical tests including repeated measurement analysis, one-way ANOVA, Kruskal Wallis H, and χ² were conducted by SPSS software (version.22). The effect of baseline differences was adjusted using ANCOVA. The post-hoc Tukey’s HSD test was conducted to confirm power analysis. Tukey’s HSD test compares the difference between each pair of surgery types group means to determine which differences are large enough to be considered significant. A p value lower than 0.05 was considered significant.

The participants of the present study were a total of 2,567 with morbid obesity who had undergone three types of bariatric surgery. The percentage of loss to follow-up was 34% until 5-year after surgery. The mean age of patients was 39.03 ± 26 which was not significantly different between operations. Most participants were women (80.1%). The mean preoperative weight and BMI were 122.93 ± 17.76 (kg) and 45.67 ± 5.47 (kg/m²), respectively. The prevalence of IGT, T2DM, hypertension, dyslipidemia, and hypothyroidism were 17.4%, 18.65%, 19.90, 40.31%, and 19.90%, respectively. Further, 17.99% and 53.79% of participants suffered from sleep apnea and low-back and knee pain. In addition, 3.23% of patients had a history of cardiovascular disease (Table 1).

The mean values of %EWL at 6 months after each procedure were comparable, and mean differences were not statistically significant (65%, 62%, and 63% EWL for OAGB, RYGB, and LSG, respectively, p = 0.468). At 9 months after surgery, OAGB represented a significantly higher %EWL than RYGB and LSG (76%, 71%, and 71% respectively, p < 0.001). RYGB and LSG did not show a significant difference in weight loss trend until 24 months after surgery. 24 months postoperatively, RYGB illustrated a significantly higher %EWL respect to LSG (77% vs. 71%, p < 0.001) (Table 2). At 3 months after surgery, OAGB represented a significantly higher %TWL than RYGB and LSG (20.83%, 19.87%, and 19.9% respectively, p < 0.001) (Table 3). The number of included participants in each time point is reported in Tables 2 and 3.

The maximum %EWL and %TWL for 5 years postoperative was obtained in 18th month after all three types of OAGB, RYGB, and LSG. The time of 18 months following surgery was the most probable time point for initiation of weight loss plateau (Table 3).

In 18 months after surgery, LSG showed significantly weight regain rather than other types of surgery. The most weight regain in three procedures were taken place in 36–48 months’ interval post-surgery. The total 5-years weight regains in three procedures (6.88%, 5.3%, and 3.3% for LSG, RYGB, and OAGB, respectively) were significantly different.

From a total of 2,567 participants of the present retrospective cohort study, the 5-year follow up rates were 85% (n = 268), 81% (n = 704), and 78% (n = 1,077) for LSG, RYGB, and OAGB; respectively. The 5-year %EWL and %TWL trend revealed several weight fluctuations from the time of surgery. This trend was significantly different among the three types of surgery (p < 0.001) (Fig. 1). The mean %EWL/%TWL, 5 years post-LSG was almost comparable to the amount of weight loss in 6 months’ post-surgery. Further, the 5-year mean weight loss after RYGB was equal to the %TWL which was obtained in 9 months following surgery. In addition, for OAGB, 5-year mean %EWL was comparable to %EWL in 12 months’ post OAGB surgery. Accordingly, the mean %WL in 6, 9, and 12 months after LSG, RYGB, and OAGB, respectively, could predict the %WL outcome 5 years’ post-surgery (Fig. 2).

The results of the present study showed that the mean %EWL in the first 6 months after surgery was independent of the type of surgery and was similar in different surgeries. OAGB and RYGB induced weight loss speed up after nine and 24 months (respectively) as compared to LSG. Weight plateau and weight regain occurred earlier and higher in LSG during 5 years. The 5 years %WL could be comparable with %WL in 6, 9, and 12 months after LSG, RYGB, and OAGB, respectively.

Based on the findings of this study, OAGB creates the fastest and highest %EWL which is followed by RYGB and LSG in the 5-year %EWL trend. The results of prospective randomized comparative studies of short- and long-term weight loss were in line with our study [9]. Their findings showed that OAGB yields a significantly greater %TWL than RYGB and LSG at 1, 2, and 5 years after surgery. However, they did not find significant %TWL differences between RYGB and LSG in the short- and long-term [10]. Based on another study to compare the effects of the LSG and RYGB based on a meta-analysis of 28 articles, the pooled analysis of 3-year and 5-year follow-up revealed that RYGB had a significant advantage over LSG for the %EWL as well as remission of T2DM. [11‒13]. So, these differences in weight reduction of various types of surgery should be considered in the patient selection and choice of suitable type of surgery [14]. It seems that a subset of bariatric patients fails to maintain their weight loss and may struggle with weight loss plateau or weight regain [15]. Several studies suggested that approximately 50% of patients may experience weight loss plateau. In addition, at least 10–30% of patients regain their weight as early as 18–24 months’ post-surgery [16, 17]. The results of the present study support the finding of previous studies that the time interval of 18–24 months after operations is the most probable time to hitting in weight loss plateau and even weight regain, especially for LSG, which induces the earliest and highest weight regain during 5-year follow-up after surgery. In LSG, as a restrictive procedure, expanding sleeved stomach caused by dietary failure or unmodified feeding behavior was the most common cause of weight loss resistance and/or weight regain. Other factors that influence weight regain after surgery include the presence of binge-eating disorders [16, 18]. Obesity comorbidities often return with weight regain [19]. Furthermore, its management is difficult with the dietary approach or lifestyle modification alone [17, 20]. Previous studies have found that among patients experiencing weight loss resistance or weight regain, 60% never adopted nutritional follow-up and 80% never passed psychological follow-up [21]. So, the prevention of weight regains and weight loss plateau are suggested as the first therapeutic step which is possible when the patient and surgeon team are aware of the common risk factors and most probable times when the weight loss speeds down and/or weight could relapse.

The 5-year %TWL after surgery was predictable by %TWL in 3–6, 6–9, and 9–12 months after LSG, RYGB, and OAGB, respectively. The present findings were comparable with the results of the Manning et al. [22] study. They showed that the percentage of weight loss 3–6 months after LSG and RYGB was an independent predictor of maximal %TWL [23]. Another prospective study of %EWL trend revealed that early weight loss trend could predict the mid-term 3-year outcome of patient underwent RYGB [24, 25]. Various factors, including gender, age, preoperative weight, in addition to the patient’s orientation, are determined to affect long-term weight loss and outcomes of bariatric surgeries [26]. Many previous studies on long-term weight loss prediction after bariatric surgery generally suggest that the status of early response to bariatric surgery could determine the long-term weight loss expectation. The present finding suggests that early postoperative weight loss can be used to predict mid-term outcomes after bariatric surgeries. It provides the opportunity for early identification of patient’s suboptimal weight loss which could allow earlier consideration of intensive postoperative lifestyle and/or behavioral modification to enhance their maximum long-term weight loss. The examples include regular and tighter nutritional and psychological follow-up visits to modify emotional eating, graze eating, irregular frequent snaking, sweet/high-calorie eating, as well as uncontrolled psychological disorders and life stressors plus lack of regular physical activity.

Some strengths of the present study included reasonably large sample size, mid-term follow-up of post-operative patients. Using data from a nationwide referral center of obesity surgery registry could assure the generalizability of our findings. Further, the comparison of outcomes between the three most popular types of surgeries contributes to the most appropriate individualized selection of types of surgeries. Retrospective design of the study could be a study limitation. Prospective multicenter large studies with longer follow-up, controlling for nutritional regimens, physical activity, weight-related medications, compliance with these regimens, and medications are suggested to be conducted in future studies.

The longer term result of surgery remains one of the most important concerns of patients and the bariatric surgery team. OAGB provides the fastest and highest %EWL/TWL, while LSG induced the earliest and largest weight plateau plus weight regain during 5 years’ post-surgery. The pattern of early weight loss could predict the mid-term outcome of bariatric surgery. So, early identification of patient’s suboptimal weight loss could allow consideration of earlier intensive postoperative lifestyle and/or behavioral modification to enhance their maximum long-term weight loss.

The present project was approved by the Ethics Committee of Iran University of Medical Sciences with the ID number of IR.IUMS.REC 1395.95-04-140-29239. Written informed consent was obtained from all individual participants included in the study.

The authors declared no conflicts of interest regarding the publication of the present article.

No funding was obtained for this study.

Mahsa Hatami: conceptualization, data collection, writing original draft, and review and editing of final draft. Abdolreza Pazouki: conceptualization, supervision, data collection, and review and editing of final draft. Fatemeh Sadat Hosseini-Baharanchi: conceptualization, writing original draft, methodology and analysis, and review and editing of final draft. Ali Kabir: conceptualization, supervision, methodology and analysis, and review and editing of final draft.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.