Bariatric surgery and mental health outcomes: an umbrella review

Aims To evaluate the breadth, depth and effectiveness of the evidence quality of all existing studies on bariatric surgery and mental health outcomes. Design Umbrella review of existing Systematic review and meta-analyses. Data sources PubMed, Embase, Web of Science, and the Cochrane Liberally databases of Systematic review and meta-analyses, and hand searching the reference lists of eligible publications. Results The search identified nine studies and 20 mental health outcomes from 1251 studies. Evidence shows that bariatric surgery is associated with significant improvement in areas such as anxiety, depression and eating disorders (including binge-eating disorder), and there is a significant harmful association with suicide, self-harm and alcohol use disorder (AUD). Among them, the most studied outcome is depression (4 articles). High-quality evidence proves that the score of depressive symptoms can be significantly improved after bariatric surgery within a two-year follow-up period and is not affected by the follow-up time. Low-quality evidence shows that bariatric surgery can significantly reduce depressive symptoms regardless of age and BMI, with an odds ratio (OR) of 0.49. Regardless of the postoperative BMI, the anxiety symptoms of women over 40 still decreased significantly, with an OR of 0.58. Regardless of the type of surgery, surgery can significantly reduce the incidence of eating disorders and symptoms. However, there is no obvious change in the follow-up time of AUD in the first two years after bariatric surgery, and the risk increases obviously in the third year, with an OR of 1.825. The evidence of moderate research shows that the risk of suicide and self-harm increases after bariatric surgery. The odds ratios in the same population and the control group were 1.9 and 3.8 times, respectively. Conclusion Bariatric surgery is beneficial for improving most mental health-related outcomes. However, we should be cautious about the increased risk of adverse mental health after surgery, such as suicide, self-harm, and AUD.


Introduction
Obesity has become one of the most severe global public health problems of the 21 st century-more than 35% of men and 40% of women with obesity in the United States (1,2).Obesity may be associated with social stigma and self-potential shame, causing a psychological burden on patients with obesity and may increase with body mass index (BMI) (3).There has been growing evidence of a two-way association between mental illness and obesity, particularly among bariatric surgery candidates (4).Depression and anxiety are common mood disorders that often coexist with obesity (5).In the general United States population, the prevalence of depression in patients with obesity seeking surgery exceeds the published prevalence (19% VS 8%) (5).Depression has long been thought to be significantly associated with binge-eating disorders (BED) (6).Eating disorder is associated with an additional burden on patients with obesity, Such as low quality of life, impulsivity, and emotional regulation disorders (7), in the regulation of hunger/ satiety (8), on the palate (9), on food preferences or intolerances (10,11), and a higher likelihood of coexisting mental health conditions.
Although bariatric surgery has been shown to alleviate complications related to physical health, its effects on mental illness have not been clearly articulated (12).In the United States, the rate of bariatric surgery remains < 1% among the eligible population (13).This low rate may be due to questions about the long-term effectiveness of bariatric surgery (14).Bariatric surgery was associated with significant improvements in depression, BED (15, 16), and eating disorders (17).While the results were encouraging, with long-term follow-up data suggesting that some postoperative patients did not experience psychological benefits or reported increased rates of depression and BED recurrence (16,18).
People are concerned about the potential risks of mental health disorders after bariatric surgery, including self-harm, suicide, and substance abuse (19,20).One study reported that patients after bariatric surgery had a 1.98-fold increased risk of suicide compared to usual care of patients with obesity (21).The incidence of suicide and suicide attempts occurs on average 3.8 to 3.9 years after surgery (22).In addition, the incidence of self-harm and alcohol use disorders (AUD) increases postoperatively (23,24) and may be associated with changes in reward mechanisms after bariatric surgery (25).we re-evaluate the meta-analysis/systematic review of all the mental health outcomes of bariatric surgery to reveal the quality and strength of the evidence.

Method Search strategy
This umbrella evaluation is completed according to the previous standardised procedures (26, 27).SL, SD, and FZ systematically searched PubMed, Embase, Web of Science, and the Cochrane Library databases until February 1, 2023.Within each theme "or" operators were used to combine terms, the "and" operator is used between different topics.Searches were tailored according to the functionality of each database.Where possible, MeSH terms were used that corresponded to the thematic areas.To avoid missing relevant meta-analyses during the initial search, we hand-searched the reference lists of eligible publications.The flowchart of the selection process is shown in Figure 1.We used 'bariatric surgery' and 'mental health' and 'meta-analysis/systematic review' and their synonym as keywords in the search database.Because mental health is disturbed by a wide range of external and internal factors, we choose the most relevant and common mental illness related to bariatric surgery as the classification of the psychological outcome, including anxiety, depression, suicidality, suicidal ideation, posttraumatic stress disorder (PTSD), substance abuse disorders, personality disorders, and eating disorders-primarily BED.No restrictions were placed on language.The detailed search strategy is in Supplementary File S1.

Eligibility criteria
The inclusion and exclusion criteria for this umbrella review are as follows: The literature included in this review is selected as follows: (1) The study reported the results of the relationship between bariatric surgery and mental health; (2) Each outcome includes at least three studies; (3) Study on the effective quantity of research report: Odds ratio (OR), Relative risk (RR), standardised mean difference (SMD), prevalence rate (PR), and Hazard Ratio (HR); (4) Aggregate effect quantity is 95% confidence interval (CI); (5) Conducted a systematic review with meta-analysis.Without adding any language restrictions, we excluded studies that could not extract data, such as guidelines, review articles, animal studies, case reports, letters, posters, conference abstracts, systematic review protocols, and other basic studies and book chapters.

Study selection
SL and FZ independently screened titles and abstracts, which decided on the full text of systematic reviews/meta-analyses that may meet the requirements.Disagreements are discussed by the third author (SD) and resolved by consensus.For any article that needs to be translated into English, online translation software (Google Translation) is used at first because it shows reasonable accuracy (28).SL recorded the research choice of systematic review according to the PRISMA statement based on systematic review and meta-analysis (29).Patient/population, intervention, comparison, outcome, and study design (PICOS) criteria are as follows (Table 1).

Data extraction
SL, SD, and FZ extracted all data independently.Differences are resolved through discussion and consensus.When meta-analyses included multiple outcomes, each outcome was extracted separately.Extract the following data from the final included article: Mental health-related outcomes, first author and publication year, follow-up, type of bariatric surgery, types of meta-analysis and subjects' number, the original study design, the metric of effect size, the effective model of meta-analysis, the effect size of 95%CI, the Heterogeneous P-value or I 2 , and publication bias.
Assessment of methodological quality SL, SD, and FZ independently used AMSTAR-2 to evaluate the methodological quality of each meta-analysis.The AMSTAR-2 tool provides a comprehensive critical evaluation tool to evaluate the systematic review of health interventions (30).AMSTAR-2 consists of 16 items, 7 of which are key areas.Each review was scored on whether there were methodological flaws in key or non-critical items.The grades are 'High', 'Moderate', 'Low' and 'Critically low'.Disagreements were resolved through discussions, although a provision had been made to consult a fourth reviewer (ZD) if necessary.
The credibility of the evidence SL, SD, and FZ independently evaluated the quality of mental health-related outcomes using the GRADE framework (31).Flowchart of the article selection process.
According to the assessment of inconsistency, risk of bias, indirectness, inaccuracy, and publication bias of each outcome, the quality of evidence is divided into four categories ('high', 'medium', 'low', and 'very low') (32).

Data analysis
The purpose of the Umbrella Review was not to perform a metaanalysis of included studies, such as assessing study eligibility or the risk of bias, but rather to use relevant details extracted from the included meta-analysis to generalise the results for specific questions.It provides the existing and highest-level evidence about medical research and draws more reliable conclusions (26).We extracted only the existing effect size and 95% confidence interval for each outcome, which provided a manageable range of effect sizes for the study.Heterogeneity is expressed in P-values or I 2 values, with P-values <0.1 or I 2 ≥50% considered significant heterogeneity.Publication bias P-values< 0.1 were statistically significant.

Corrected covered area index
When several meta-analyses investigated the same outcome, we conducted a corrected covered area index analysis on the repeated included literature.The degree of overlap in studies was assessed and calculated via the corrected covered area (CCA) index method (33).CCA within the range 0%-5% express a slight overlap, 6%-10% express a moderate overlap, 11%-15% express a high overlap, and >15% express a very high overlap (34) (Table 2).

Result Literature review
A total of 1251 potentially eligible articles were identified: 117 from PubMed, 188 from Web of Science, 792 from Embase, and 20 from Cochrane Library.We screened the titles of 1251 potential studies, excluding 170 duplicate articles and 954 studies not related to studies.We read abstracts, assessed the eligibility of the remaining 127 studies, excluded 88 studies, and classified the studies according to reasons for exclusion.Finally, nine eligible studies were selected by reading the full text, and 30 other studies that did not meet the requirements were excluded.The flow chart of the selection process is shown in Figure 1.Finally, nine studies with 20 different mental health-related outcomes were included in the umbrella review.Figure 2 lists the main mental health outcomes after bariatric surgery.The relationship between bariatric surgery and mental health-related outcomes is shown in Table 3.

General mental health
For general mental health, a research report that a significant improvement could occur after an intragastric balloon (IGB) (OR = 0.86; 95% CI: 0.29 to 1.42) at six to 76 months of follow-up (35) (Figure 3).

Eating disorders
The overall postoperative prevalence of eating disorders is 7.83% (95% CI: 4.30 to 11.37).Considering the six studies that make up these, the prevalence of BED was 3.81% (17) (Figure 3).

AUD
Bariatric surgery had no significant effect on AUD after 1 and 2 years of follow-up; However, the incidence of AUD was significantly higher after three years of follow-up (OR = 1.83; 95% CI: 1.53 to 2.178) (42) (Figure 3).

AMSTAR-2 quality assessment
The methodological quality of the nine studies assessed was using the AMSTAR-2 tool.Most of the included reviews have weaknesses.Four studies (44.4%) were rated as critically low, three studies (33.3%) were rated as low, one study (11.1%) was rated as moderate, and one study (11.1%) was rated as high (Table 4).

GRADE classification
The GRADE system was used to evaluate the evidence quality of each postoperative mental health outcome.Only one outcome was rated as high quality (5.0%), two (10.0%) as moderate, thirteen (65.0%) as low, and four (20.0%) as very low (Table 5).

Discussion
This umbrella review summarised the highest-level evidence on bariatric surgery for multiple mental health outcomes.We included nine meta-analyses/systemic reviews assessing 20 different mental health outcomes.This study found that bariatric surgery was beneficial to improving overall mental health, reducing depression and anxiety symptoms in patients with obesity, and eating disorders (e.g., BED) also improved after surgery.In addition, bariatric surgery can increase the risk of suicide, self-harm, and increased AUD after surgery.Although the mental state has improved after the operation, the mechanism behind this evidence is still unclear.It may not be bariatric surgery itself, but the life changes caused by weight loss (caused by bariatric surgery).The same is true of harmful consequences.
Bariatric surgery can significantly improve the symptoms of depression and anxiety and reduce the incidence; this has been proved in many meta-analyses (35,36,38,39).
As expected, The degree of change in depressive symptoms is related to the shift in BMI after the operation (43).Although bariatric surgery can improve depressive symptoms in the short term, long-term follow-up studies show that the Hospital anxiety and depression scale is equivalent to before surgery in four years and nine years after bariatric surgery (44, 45).There may be several paths in its mechanism.(a) First, body image satisfaction, selfworth, interpersonal relationships (5), and an increase in physical activity can be improved after the operation (46,47).Secondly, after Forest plot of mental health outcomes associated with bariatric surgery.(48).Third, patients whose depression is likely to improve on their own are more likely to undergo surgery.Notably, studies have reported an 8% to 74% reduction in the prevalence of postoperative depression (5).However, the prevalence of postoperative depression is still as high as 15.3%, of which mild depression accounts for 64.9% (37); that may be related to weight regain following bariatric surgery, eating disorders, and impaired quality of life.
There was a significant improvement in BED (49, 50) after bariatric surgery, especially in the first 12 months after surgery, and larger effect sizes were found.One study included seven studies that reported the prevalence of eating disorders as 7.83% (17).Even if the postoperative prevalence is less than 10%, the disease can significantly affect prognosis and BMI reduction.Among them, the prevalence of BED is 3.81% (17).After bariatric surgery, avoiding high-fat/high-sugar foods may be a learned response to postprandial discomfort or dumping syndrome (51), this in turn helps to control obesity in this population (52).With longer postoperative follow-ups, conflicting outcomes emerged.Smith et al. (53) reported that BED increased from 2.1% to 4% after seven years, and loss of control eating increased from 24.6% to 26.4%.Conversely, Kalarchian et al. (54) report that the prevalence of eating disorders drops to 0% within seven years.The reason may be that preoperative comprehensive nutrition and psychological behaviour evaluation, education, and support decreased with time (55).
Notably, bariatric surgery is associated with an increased risk of suicide, self-harm, and AUD (40,42).The prevalence of suicide after bariatric surgery was 0.3% (41).Peterhansel et al. (19) who reviewed suicide after bariatric surgery, reported that suicide rates were estimated at 4.1 per 10,000 people per year, with aboveaverage suicide rates in bariatric surgery patients.Lack of improvement in quality of life after surgery, persistent or recurrent sexual dysfunction and interpersonal problems, and limited physical activity can all lead to an increased risk of suicide (20).The degree of BMI reduction does not seem to explain the change in mental distress after bariatric surgery (56).Compared with other patients undergoing surgery, the patients who have committed suicide or attempted suicide have similar or more significant BMI reduction (57).People with pre-existing mental illnesses such as depression and eating disorders are more likely to commit suicide after surgery due to the underlying psychiatric effects of people with obesity (58).The surgery itself increases potential postoperative problems, such as difficulty controlling pain leading to substance abuse, complications requiring further treatment, dissatisfaction with BMI reduction, excess skin, and scarring (59)(60)(61).Neuroendocrine changes and nutritional deficiency caused by postoperative malabsorption and/or irregular eating behaviour may also increase the risk of suicide (62).Because of the potential symptom change after the operation (cross-   addiction hypothesis: from overeating to excessive drinking), it may be related to depressive symptoms and suicide attempts, which leads to an increase in the risk of AUD (63,64).Bariatric surgery is associated with a significant increase in moderate to high-risk AUD (42).One study reported an 8% incidence of postoperative alcohol use/abuse; patients were more likely to drink heavily after surgery than before surgery, 19% had significant alcohol consumption before bariatric surgery, and 23% had apparent alcoholism after bariatric surgery (65).Of the three studies that reported rates of alcohol abuse (23,66,67), two studies reported an increase in alcohol problems after RYGB (23, 67), another study reported that no alcohol problems were found in LAGB (66).Increased alcohol consumption after RYGB may be due to physiological changes after gut bypass that are absent after LAGB (including increased release of glucagon-like peptide 1 and peptide YY after meals and decreased circulating acyl-ghrelin) (68, 69); these changes contribute to changes in neural activity (70).Intestinal flora may affect mood and behaviour, and exposure to abused drugs, such as alcohol and opioids, will induce intestinal microbial ecological imbalance (48).
In addition, we found no evidence to support a link between personality disorders and PTSD and changes in BMI reduction outcomes.Depression and anxiety are common among candidates for bariatric surgery.About 40% of patients with obesity have underlying psychiatric disorders, and early identification and optimisation of these disorders are important because they can affect surgical outcomes (71).In the future, we still need to pay attention to the mental health of patients with obesity and assess their mental health status before and after surgery.

Limitations
Our umbrella review was limited to evidence included in the review search window, creating limitations regarding study timeliness.Those with more severe depressive symptoms may not be eligible for bariatric surgery or cannot participate in studies and/ or stop participating in longitudinal studies (72, 73).It is unclear whether the increased risk of mental health problems after bariatric surgery is due to the surgery or the factors that force them to seek surgical treatment.The relationship between bariatric interventions and psychiatric outcomes is likely to be confounded by psychosocial features, limiting the value of inferences drawn from observational studies.Due to a lack of relevant data, we did not separately analyse the effects of different bariatric surgeries on mental health, and different BMI reduction methods were associated with other mental health outcomes.We should investigate the relationship between different BMI reduction methods and mental health-related outcomes in the future.In addition, the methodological quality of the meta-analysis we included is not high to the evidence of the outcome and stratified results.There is a need to track bariatric surgery's mental health-related outcomes in real-time and summarise the latest evidence.

Conclusion
We found that bariatric surgery effectively improved mental health and most mental health-related outcomes, such as anxiety symptoms, depression symptoms, eating disorders, and BED.However, these should be cautiously approached after bariatric surgery due to the increased risk of unfavourable mental suicide, self-harm, and AUD.

TABLE 1
PICOS criteria used to define the research question.
suicidality or suicidal ideation, and eating disorders-primarily binge eating disorderStudy DesignConducted as a systematic review with meta-analysis and included mental health outcomes FIGURE 1

TABLE 2
The overlapping included systematic reviews and meta-analyses.
FIGURE 2Map of mental health outcomes associated with bariatric surgery.

TABLE 3
Relationship between bariatric surgery and mental health-related results.

TABLE 4
Methodological quality assessment of included meta-analyses according to the AMSTAR-2.

TABLE 5
GRADE quality assessment., changes in digestion or intestinal absorption may change biochemical signals in the brain; for example, changes in intestinal microflora, intestinal peptides (including ghrelin, glucagon-like peptide 1, peptide YY and cholecystokinin) are related to depression and anxiety, and bile acids can have a central impact on the emotional and behavioural responses caused by cocaine surgery