Introduction

Obesity is a growing national epidemic that currently affects 40% of adults in the USA.1 Globally, obesity accounts for approximately 4 million deaths and represents 7.1% of all-cause mortality.2 Multiple disease processes are closely associated with obesity, including high blood pressure, high cholesterol, and type 2 diabetes mellitus. These conditions not only contribute to the comorbidity burden in obese patients, but also serve as major risk factors for cardiovascular disease.3, 4 Additionally, obesity is associated with a reduced health-related quality-of-life, functional limitations, and poor mental health.5

Bariatric surgery has been shown to be a highly effective long-term weight loss intervention with laparoscopic Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG) being the most common procedures employed.6,7,8 In addition to facilitating weight loss, a meta-analysis by Buchwald et al. in 2005 reported that bariatric surgical interventions may also contribute to a resolution of type 2 diabetes, hypertension, and sleep apnea in a majority of patients with these comorbidities.9 Moreover, due to advances in laparoscopic techniques, bariatric surgery has become a relatively safe procedure, with a mortality rate of less than 1% across studies.10, 11 However, despite the clear benefits of surgery, previous studies have shown that less than 1% of patients who are clinically eligible to undergo bariatric surgery do so12 Although the reasons for this are not completely understood, a possible contributing factor is the requirement for patients to adhere to strict weight loss programs or meet specific weight loss criteria before qualifying for surgery.13, 14

Preoperative weight loss (PreWL) has previously been studied as a predictive factor for success of postoperative outcomes after bariatric surgery.15,16,17 While the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program recommends PreWL as a tool for optimizing comorbidities in certain patient populations, these recommendations are based on limited evidence suggesting that increased PreWL may result in a greater reduction in surgical complications.18 For instance, it has been suggested that laparoscopic surgery may be more challenging in an obese patient due to limited visualization of the abdomen and difficulty in manipulation of surgical instruments. Obese patients also have a greater prevalence of enlarged nonalcoholic fatty liver, which may contribute to increased conversions to open surgery.19, 20 Some proponents of PreWL have asserted that it may decrease liver size and lead to a reduction in surgical complications, operative time, and morbidity in patients undergoing bariatric surgery.

The existing literature has shown mixed results on the effects of PreWL on postoperative weight loss outcomes and complication rates. A recent review by Tewksbury et al. found that PreWL may not be the strongest predictor of postoperative weight loss.21 In addition, the variability of results from published studies makes it challenging to clearly recommend for or against the use of PreWL requirements for surgical eligibility. This is supported by a recent review by Kim et al., who concluded that there is insufficient evidence regarding the benefits of mandatory PreWL on postsurgical outcomes.13 Therefore, the practice of mandating preoperative weight loss prior to bariatric surgery is controversial according to available literature and may be a potential obstacle for patients requiring surgical treatment for obesity.

While most patients experience significant and persistent weight loss after bariatric surgery, as many as 20% of patients may experience postoperative weight regain.22 Several studies have attempted to correlate preoperative characteristics with bariatric surgical outcomes, but the factors associated with long-term bariatric surgical success remain unclear. The objective of this study is to evaluate the impact of PreWL on perioperative complication rates and postoperative weight loss (PostWL) for patients who underwent bariatric surgery for obesity at our institution.

Materials and Methods

This was a single-center, multi-surgeon, retrospective study of patients who underwent primary laparoscopic sleeve gastrectomy (LSG) or primary laparoscopic Roux-en-Y gastric bypass (LRYGB) surgery for obesity at our institution between January 2003 and November 2017. Data collection utilized chart review and a postoperative standardized patient questionnaire for follow-up. A total of 54.9% (234/426) of postoperative follow-up weight were obtained by telephone follow-up. This study was approved by the Institutional Review Board of the University of Southern California.

Date of initial evaluation for bariatric surgery and operative date were used to calculate length of bariatric program (LOBP). Preoperative weight loss (PreWL) was calculated using preoperative excess weight and initial evaluation excess weight. Only patients with a LOBP of 90 days or greater were included in our analysis. Height was used to calculate ideal BMI and subsequently excess weight. PostWL success was defined as greater than 50% excess weight loss (%EWL) at least 1 year after surgery.

Statistical Analysis

Categorical variables are reported as frequencies and percentages, while continuous variables are reported as means with 95% confidence intervals. Patients who underwent any additional procedures during their surgery were excluded from analysis regarding perioperative complications. Bivariate analysis was used to compare between patients who achieved EWL greater than 10% and patients who achieved EWL less than 10% for each surgery. Analysis included Chi-square and Fisher’s exact test for categorical variables and Student’s t test for continuous variables.

Multivariable logistical regression models were constructed to examine PreWL as a predictive factor for PostWL. Patients were organized into three groups based on preWL during bariatric program: (1) patients who achieved less than 5 %EWL (%EWL < 5); (2) patients who achieve greater than 5 %EWL but less than 10 %EWL (5 < %EWL < 10); and (3) patients who achieve greater than 10 %EWL (%EWL > 10) similar to a previously published report.15 Covariates included in the regression model were age, sex, surgery type, preoperative BMI, length of postoperative follow-up, and LOBP. Covariates were chosen based on clinical relevance and maximizing model fit. Univariate logistical regression was first used to determine the unadjusted effect of each covariate. A multivariable logistical regression model which included all variables to examine how these predictors changed when adjusting for covariates was then constructed. Linearity of the continuous variables with respect to the logit of the dependent variable was assessed via the Box-Tidwell (1962) procedure, and all continuous variables were found to be linearly correlated to the logit of the dependent variable. All models were statistically significant. Model fit was assessed using Hosmer and Lemeshow test. For all statistical analyses, a p value of less than 0.05 was considered significant. All analysis was performed using IBM SPSS statistics 24 software.

Results

A total of 426 patients were included in our analysis. The majority was female (n = 312, 73.2%) and underwent LSG (n = 201, 61.0%). The average age at surgery was 45.6 years, and the average length of follow-up was 6.3 years. The most common comorbidities included hypertension (54.5%), GERD (48.5%), and osteoarthritis (46.2%). Summary of patient demographics, past medical history, and past surgical history is shown in Table 1. Patients presented with an average BMI of 47.3 at initial evaluation for bariatric surgery. The average LOBP was 32.7 weeks, during which patients achieved an average of 9.7 %EWL. When categorized by PreWL during LOBP, 30% of patient achieved less than 5% %EWL, 16% achieved greater than 5% but less than 10%, and 54% achieved greater than 10% %EWL as shown in Fig. 1. Postoperatively, patients lost an average of 57.9 %EWL, and 59.8% achieved PostWL success. Summary of preoperative and postoperative data is shown in Table 2.

Table 1 Summary of patient demographics, past medical history, and past surgical history
Full size table
Fig. 1
figure 1

Distribution of patients according to preoperative percent excess weight loss (%EWL) during bariatric program

Full size image
Table 2 Summary of preoperative and postoperative data
Full size table

Perioperative complication rates for patients who achieved less than and greater than 10 %EWL are compared (Table 3). Intraoperative complications, estimated blood loss, transfusion requirement, operative time, postoperative complications, and ICU admissions were not significantly associated with PreWL for either surgery type. Greater than 10 %EWL was associated with lower length of stay (LOS) (1.3 vs 1.8) in patients who underwent SG but not RYGB.

Table 3 Comparison of the differences in demographics, past medical history, past surgical history, and perioperative complication rates for patients who achieved less than or more than 10% excess weight loss (%EWL) preoperatively
Full size table

We evaluated the association between PreWL and PostWL success. Univariate analysis revealed patients who achieved greater than 5%, and less than 10 %EWL preoperatively were O.R. 0.41 (95%CI: 0.20–0.86) as likely to achieve PostWL compared to patients who achieved less than 5 %EWL. Patients who achieved greater than 10 %EWL preoperatively were O.R. 0.54 (95%CI: 0.31–0.94) times as likely to achieve PostWL as patients who achieved less than 5 %EWL. A multivariableregression analysis was then conducted in order to adjust for sex, age, surgery type, length of postoperative follow-up, length of bariatric program, and preoperative BMI. We find no significant association between PreWL and PostWL success (Table 4).

Table 4 Univariate and multivariablelogistical regression analysis examining the association between PreWL and PostWLS
Full size table

Discussion

Bariatric surgery has been proven to be an effective modality for long-term weight loss and comorbidity resolution. PreWL is commonly mandated for patients prior to bariatric surgery despite the lack of consensus in the literature regarding its benefits. We examined the association of PreWL with perioperative complications and PostWL after bariatric surgery. We conducted a rigorous analysis and fairly stringent data requirement for our cohort including at least 90 days of LOBP for all patients. Analysis of PreWL association with PostWL included patients with at least 1-year follow-up, which resulted in a relatively long average follow-up (6.3 years) compared to other studies and analysis utilized a multivariable logistical regression model to control for possible confounding variables. Additionally, we excluded patients with concomitant procedures during surgery for analysis of complications in order to limit operative variability. We found no significant association between PreWL and perioperative complications or PostWL.

While it may be intuitive that patients who lose weight prior to surgery will likely be more successful in achieving postoperative weight loss, data from published studies on this topic have been conflicting and inconsistent. Some reports demonstrate a positive association,23,24,25 while others find no association.15, 26,27,28,29 In 2015, Gerber et al. published a study that summarized data from two randomized controlled trials, 21 prospective or retrospective studies, and reviews by Livhitset al. and Oncher et al. to comprehensively study the effects PreWL on surgical complications and PostWL.30,31,32 The authors concluded that due to the inconsistency of available data, the results cannot be used as justification for mandatory weight loss programs prior to bariatric surgery.30 Our findings echo the conclusions of this review as our multivariable regression model showed no association between PreWL and PostWL. Our findings also highlight an important implication regarding statistical analysis as we found a significant correlation between PreWL and PostWL on univariate analysis as shown in Table 4. When controlling for various factors that may influence postoperative weight loss such as age, sex, preoperative BMI, and time from initial evaluation to surgery, that difference was no longer significant. These results highlight the need for strong statistical analysis such as regressions in future studies for the examination of this relationship as PostWL is a complex outcome that may be influenced by many factors.

PreWL has also been studied as a predictive factor for surgical complications. Intraoperative and postoperative complications following bariatrics surgery include anastomotic leaks, gastrointestinal bleeding, marginal ulcers, intestinal obstruction, anastomotic strictures, and deep venous thrombosis.33,34,35 It is hypothesized that PreWL may help to reduce the frequency of intraoperative complications by decreasing liver size, visceral fat, and postoperative complication rates.36 However, current literature on postoperative complications has mixed findings with some studies showing no difference in complication rates,14, 23, 24, 31, 37, 38 while others have identified a decreased overall complication rate in those with greater PreWL.16, 39 Similarly, there is no correlation between PreWL and intraoperative complication rates, excess blood loss, or required transfusions, in either RYGB or SG patients. While the current inconsistency in the literature does not warrant mandatory PreWL for bariatric surgery, clinicians are encouraged to counsel patients regarding weight loss prior to surgery in order to optimize preexisting conditions that are common in this patient population.18

We found no association between PreWL and length of operative time. Our analysis excluded all patients who underwent additional procedures during surgery as that may significantly increase variability in operative time. A randomized clinical trial consisting of 61 patients by Alami et al. found decreased operative time with increased PreWL.23 A relatively larger randomized multi-center clinical trial consisting of 294 patients conducted by Van Nieuwenhove found no association.40 The authors suggest that this lack of difference may be due to differences in surgical expertise. More experienced surgeons in laparoscopic bariatric surgery are able to perform the procedures faster regardless of limited visibility of the abdominal landmarks. This hypothesis is applicable to our institution and could be an explanation for the lack of significance between PreWL and operative time.

LOS is a common parameter used as an indirect measure of surgical complications and morbidity. Some studies have reported a decreased LOS after greater PreWL,37, 41 including one study which showed lower length of stay in patients who lost 5 to 10% of excess body weight in high-risk obese patients.39 Other studies, including a randomized clinical trial by Alami et al., where patients were randomized to a 10% weight loss requirement or no weight loss requirement, found no difference in LOS.14, 20, 23, 42, 43 Our data shows a significant association between increased PreWL and decreased LOS in patients who underwent SG but not RYGB. While statistically significant, the difference between 1.8 and 1.3 days may be clinically insignificant.

Due to the lack of consensus regarding the efficacy of PreWL as a predictive factor of PostWL and complication rates, further studies are necessary to make definitive conclusions. A study by Sadhasivam and colleagues found that a large percentage of patients at their institution were denied surgery despite being deemed good surgical candidates.44 They found that failure to achieve substantial weight loss prior to surgery to be one of the reasons insurance companies denied coverage of bariatric surgery to eligible surgical candidates. While PreWL may be a positive and productive recommendation for patients with obesity, the existing data do not support withholding surgical intervention purely for failure to adequately lose weight prior to surgery. Insurers who mandate that patients participate in medical weight management programs without providing sufficient coverage for these services may be limiting access rather than improving patient outcomes. Clarifying the role of PreWL in bariatric surgery is essential not only for identifying potentially modifiable factors that can maximize postoperative success, but also for guiding clinicians and insurers in providing access to this highly effective weight loss treatment.

Our study has limitations inherent to its design being a retrospective single-center study. In addition, patient weight obtained through telephone follow-up is subject to external bias that may limit the strength of the conclusions derived from our analysis, such as social desirability bias. In order to potentially reduce risk of bias, patients were provided with a mission statement prior to every survey addressing judgment, anonymity, and the study objectives. Nevertheless, the risk for bias must be considered. The lack of association of preoperative weight loss and perioperative complications in our analysis may be due to insufficient power of our study and relatively low rates of complications. Nonetheless, we included this analysis in order to comprehensively understand the utility of preoperative weight loss. A major strength is that our study presents data on long-term postoperative outcomes (mean 6.3 years) compared to the literature. In addition, regression analysis provides a strong statistical tool for evaluating predictors of postoperative weight loss while controlling for potential confounding factors.

Future studies are needed to better understand the effect of preoperative weight loss on perioperative complication rates and its utility as a predictive tool for postoperative weight loss success. This can be achieved by conducting large prospective studies that are powered to detect differences in operative complications. Additionally, standardization in reporting postoperative weight loss would prove useful for performing future meta-analysis. Given the requirement of preoperative weight loss serves as an obstacle for some patients in obtaining access to bariatric surgery, more studies are essential.

Conclusions

Preoperative weight loss is not predictive of postoperative weight loss success after bariatric surgery. Greater preoperative weight loss was associated with a mild decreased in length of stay but was not associated with a reduction in operative time, overall complication rates, ICU admissions, or intraoperative complications. The inconclusive literature and our findings do not support the medical necessity of weight loss prior to bariatric surgery for the purpose of reducing surgical complications or predicting successful postoperative weight loss success.