Introduction

Obesity is considered as a disorder of the energy homeostasis system which is the consequence of complex interactions between genetic, epigenetic, developmental and environmental influences [1]. Hemmingsson recently proposed a new causal conceptual model linking socioeconomic disadvantage, psychological and emotional stress factors experienced by parents and children, and possible disrupted energy balance homeostasis, resulting in weight gain and ultimately obesity [2]. As a result, adult obesity can be considered as a delayed consequence in some individuals not only of a disharmonious childhood family environment [2], but also occurrence of adverse experiences (physical, emotional and sexual abuse, physical and emotional neglect) [3, 4].

A meta-analysis has demonstrated a positive association between childhood maltreatment and its severity with the risk of obesity in adults [5]. Incidence and severity of childhood trauma were also significantly associated with onset of psychological disorders (depression, anxiety, post-traumatic stress disorder, alcohol addiction, suicide ideation, binge eating) [3]. The prevalence of emotional neglect, physical, emotional and sexual abuse is high in patients with severe obesity [5], especially in those with binge eating disorder (BED) [4, 6, 7]. Up to 50% of patients seeking bariatric surgery have a history of BED [8], which can be considered both as one of the major psychopathological symptoms in patients with severe obesity as well as a marker of general vulnerability during the life course.

Protective factors against psychological and emotional distress, particularly self-esteem and resilience, play a crucial role in the Hemmingsson model [2]. Despite the increasing use of the term « resilience», a literature review underscores the lack of universal definition [9]. Thus, resilience is a process allowing to achieve positive outcomes, successful adaptation, and leading to a trajectory of positive development despite exposure to significant risks of adversity [10,11,12]. It provides an integrative model, bringing together internal (individual capacity) and external (interpersonal resources) factors that facilitate adaptation to adversity over time [13]. In summary, resilience can be conceptualised as a process that influences and moderates the relationship between risk factors and outcome variables [14]. As a mechanism of protection, resilience increases the chance of adaptive responses and prevents the development of negative mental health outcomes [15]. In the general population, low resilience status appears to be associated with a higher body mass index (BMI) [16,17,18]. However, few studies have examined resilience in patients with obesity, and have moreover yielded conflicting results depending on the methodology used [17]. In children, obesity and overweight have been associated with exposure to adverse family experiences and the child’s lack of resilience [19]. This lack of resilience could also explain the occurrence of eating disorder during stressful situations [20, 21]. Conversely, resilience factors predict improvement in quality of life, such as psychological health and social relationship, as well as a reduction in eating disorder symptoms over time [22]. Two studies assessing resilience in the context of bariatric surgery have recently been published. For the first time, associations between pre- and post-operative psychological outcomes and resilience after bariatric surgery were investigated by McGarrity et al. using a well-validated resilience scale. Resilience scores were significantly associated with improvements in symptoms of psychological disorders from pre- to post-operative assessments, especially for binge eating symptoms, depressive and anxiety symptoms, but also approach coping scores [23]. Positive associations between coping strategies and the resilience of bariatric surgery candidates were also reported in a study by Bozdogan et al. [24].

No study to date has focused on the link between resilience capacity, adverse childhood experiences (ACE) and psychological disorders in patients with severe obesity, particularly in those seeking bariatric surgery. In this particular population, low socioeconomic status (SES) [25] on the one hand, and BED or psychiatric disorders (anxiety, depression) on the other, are highly prevalent [26]. In light of the above, the aim of the present study was to investigate the relationship between ACE, resilience, and current psychological disorders. Also examined was whether resilience play a mediation role in the association between ACE and psychological disorders in a cohort of patients seeking bariatric surgery.

Methods

Participants

All patients with severe obesity scheduled for bariatric surgery at the Nancy University Hospital (multidisciplinary bariatric surgery unit) between September 2016 and April 2017 were included in this study. During this period, 204 patients were interviewed by the psychologist. Among them, 4 patients were opposed to using their data for research, thus 200 patients were finally included in this study. All fulfilled the criteria for bariatric surgery indications defined by the French National Authority for Health [27] and European guidelines [28], which include a BMI of at least 40 kg/m2 or of at least 35 kg/m2 with comorbid conditions, associated with failure of previous conservative attempts to lose weight. To be included, all patients were required to have completed the Resilience Scale for Adults (RSA) questionnaire and also have attended a semi-structured clinical interview conducted by the same psychologist to investigate the prevalence of psychological disorder and ACE. Patients completed the French version of the RSA questionnaire [29] prior to the interview with the psychologist. All subjects were included after ensuring that they had not objected to the use of the data. All procedures contributing to this observational retrospective study (non-interventional) were in compliance with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. Approval was obtained from the Nancy University Hospital Ethics Committee (reference: 2020PI203). Data analysis was registered in the ClinicalTrials.gov (identifier NCT05341414).

Structured clinical interview

Structured clinical interview for the diagnosis of current psychological disorder

The Mini International Neuropsychiatric Interview 5.0.0 (MINI-Plus, French version) is a structured and standardised diagnostic interview used to determine the most common psychiatric disorders, according to the DSM-IV-TR Axis I Disorders and the International Classification of Diseases and Related Health Problems (ICD-10) [30]. The MINI-Plus features good psychometric properties, widely used in psychiatry, and was conducted by a trained psychologist for the study. The following most frequent diagnoses for patients with severe obesity were selected: (1) Depression, (2) Generalised anxiety disorder, (3) Substance use (alcohol, tobacco, or drugs), (4) Post-traumatic Stress Disorder (PTSD), and (5) Binge Eating Disorder (BED) [31]. The first four were diagnosed by the psychologist using the MINI-Plus, while BED was diagnosed according to the DSM-5 diagnostic criteria [32].

The total score corresponds to the number of psychological disorders diagnosed among depression, generalised anxiety disorder, substance use, PTSD, and BED (0–5).

Structured clinical interview for adverse childhood experiences

The psychologist conducted a structured clinical interview to identify ACE. For each patient, based on the Structured Clinical Interview for DSM-IV Childhood Diagnoses [33], the psychologist recorded, if relevant, the type of trauma or adversity: emotional and physical neglect, sexual abuse, physical abuse, emotional abuse. Emotional neglect was defined as the failure of caretakers to meet basic emotional and psychological needs (including love, belonging, nurturance, and support). Physical neglect was defined as the failure of caretakers to provide for basic physical needs (including food, shelter, clothing, safety, and health care). Sexual abuse was defined as sexual contact or sexual conduct forced upon the patient without consent. Physical abuse was defined by bodily assaults directed toward the patient that posed a risk of or resulted in injury. Emotional abuse was defined as verbal assaults on sense of worth or well-being or any humiliating or demeaning behaviour repeatedly directed toward the patient. A final score was calculated by adding these ACE types for each patient. This sum corresponds to the “cumulative adverse childhood experiences” score (0–5).

Resilience scale for adults (RSA)

RSA is a performance questionnaire [34], with a stable scale [15, 35] and a high sensitivity to clinical change [34], and was translated and validated in French [29]. RSA includes an evaluation of both family and social protective factors of resilience [36]. RSA is a 33-item self-report instrument, where each item-response ranges from 1 to 7, with higher scores reflecting higher levels of protective factors of resilience [37]. The RSA evaluates six protective components of resilience in adults, namely:

  1. 1.

    Perception of the self (e.g.: my personal problems, are unsolvable/I know how to solve) (six items)

  2. 2.

    Planned future (e.g.: I feel that my future looks, very promising/uncertain) (four items)

  3. 3.

    Social competence (e.g.: Meeting new people, is difficult for me/something I am good at) (six items)

  4. 4.

    Structured style (e.g.: I am good at, organising my time/wasting my time) (four items)

  5. 5.

    Family cohesion (e.g.: My family is characterised by, disconnection/healthy coherence) (six items)

  6. 6.

    Social resources (e.g.: The bonds among my friends are weak/strong) (seven items)

These six components represent three dimensions [15]:

  • “Individual dispositions”: factors 1, 2, 3 and 4 (20 items)

  • “Family coherence”: factor 5 (six items)

  • “Social support”: factor 6 (seven items)

Total score is the mean of all items, without weighting.

Other measurements

Data on height, weight, onset of weight gain, marital status, employment status, sex, and age were available in the patients’ medical files. BMI was calculated by dividing weight in kilograms by the square of the height in meters. The onset of weight gain included two categories: childhood / adolescence (< 18 years) and adulthood. Marital status was defined as follows: currently married/cohabiting, formerly married (divorced/separated/widowed), never married. Professional status consisted of five categories: active, unemployed, homemaker, retired, and unable to work.

Statistical analyses

All descriptive and inferential statistics were performed with SPSS version 25.0 [38]. Chi-square tests were applied to determine the presence of sex effects for variables pertaining to prevalence of adverse psychological events and psychological disorder. Asymmetry and kurtosis were calculated to evaluate the normality of the distribution. All quantitative parameters were between −2 and + 2, validating the normality of RSA distribution. The Mann–Whitney test was used to verify sex effects, comparing mean differences for age, BMI, number of psychological disorders, cumulative ACE, and for RSA scores. The Mann–Whitney test was also used to compare the differences in RSA according to ACE and psychological disorders. Spearman’s rho correlations were calculated between resilience, age, BMI, cumulative ACE, and number of psychological disorders. A multivariate analysis was used to explain the RSA scores according to two analysis models. The first (model A) explains three dimensions and total RSA score according to age, sex, BMI, cumulative ACE, and number of psychological disorders. The second analysis (model B) explains RSA scores according to age, BMI, sex, types of ACE (emotional, physical and sexual abuse, and emotional and physical neglect), and psychological disorders (depression, anxiety, substance use, PTSD and BED). Significance level was set at 0.05. Finally, several mediation analyses were performed with the SPSS PROCESS version 3.5 macro-extension [39], applying Model 4. Resilience (RSA total score), individual disposition, family coherence, and social support were used as potential mediators to determine if these factors can mediate the relationship between cumulative ACE (independent variable) and number of psychological disorders (dependent variable). Each analysis utilised 5000 bootstrap re-samples, and significance was determined based on 99% bias-corrected confidence intervals. Sex, age and BMI were covariates and were included in each analysis.

Results

Characteristics of the study participants

This study included 200 patients (47 men and 153 women) aged 43.5 ± 11.8 years old with a mean BMI of 45.5 ± 7.2 kg/m2 (Table 1). The onset of weight gain occurred predominantly before adulthood (under the age of 18) (64%). The majority of the subjects were currently married or in cohabitation (68%) and were professionally active (66%) (Table 1). There was no significant difference between men and women in terms of BMI, age, weight gain onset, familial status, and professional status.

Table 1 Demographics and characteristics of the 200 bariatric surgery candidates
Full size table

Resilience

The mean total RSA score was 5.16 ± 0.87 (Table 1). Men had a higher total score than women (5.32 ± 0.85 vs. 5.11 ± 0.87; p = 0.036). The mean scores for individual dispositions, family coherence and social support were, respectively, 4.93 ± 0.94, 5.26 ± 1.24 and 5.73 ± 1.08, with no significant difference between men and women. The three dimensions of RSA were highly correlated with the total score: individual dispositions R = 0.90 (p < 0.0001), family coherence R = 0.746 (p < 0.0001), and social support R = 0.801 (p < 0.0001). The overall internal consistency of the RSA measured by Cronbach’s alpha coefficient reached 0.92.

Adverse childhood experiences

Among ACE, the most frequent were emotional neglect (90.5%) and emotional abuse (61%) (Table 1). Overall, 96% (n = 192) of the patients reported at least one ACE. The mean of cumulative ACE was 2.15 ± 1.15 and 70.5% had at least two types of ACE. Overall, there was no significant difference according to sex, except for sexual abuse (women: 16.3% vs. men: 0%; p = 0.003).

Current psychological disorders

Overall, 69.5% (n = 139) of our population presented at least one current psychological disorder, without significant difference according to sex (women: 70.6% vs. men: 66.0%; p = 0.546) (Table 1). Among women, the most frequent disorders were anxiety (40.5%) and BED (34%), while in men, the most frequent were substance use (40.4%) and BED (38.3%). The prevalence of depression (15%) and PTSD (11.5%) was relatively low in both sexes. The mean number of current psychological disorders was 1.28 ± 1.16, with no significant sex differences among all current psychological disorders, except for anxiety (women: 40.5% vs. men: 21.3%; p = 0.016).

Relationship between clinical parameters and resilience

The total RSA score was significantly lower in instances of sexual abuse (p < 0.0001), emotional abuse (p = 0.001), and physical neglect (p = 0.012) (Table 2). Individual dispositions and social support dimension scores were significantly lower in instances of sexual and emotional abuse, and physical neglect, while RSA family coherence score was significantly lower in instances of sexual and emotional abuse.

Table 2 Total and RSA dimension scores (means + / − SD) according to presence or absence of adverse childhood experiences and psychological disorders
Full size table

The total RSA score was significantly lower in patients with psychological disorders such as anxiety (p = 0.001), depression (p = 0.002), substance use (p = 0.013), and BED (p = 0.002) (Table 2). Among all RSA dimensions, individual disposition was the most responsive to the presence of psychological disorders. Indeed, the individual disposition scores were significantly lower in the presence of depression (p < 0.0001), anxiety (p < 0.0001), addiction (p = 0.004), PTSD (p = 0.011), and BED (p = 0.0004). Regarding the other RSA dimensions, only the presence of substance use was significantly associated with the family coherence score (Table 2).

Respective correlations were subsequently analysed between total RSA score and each RSA dimension as well as with age, BMI, cumulative ACE, and number of current psychological disorders (Table 3). The number of current psychological disorders was significantly and negatively correlated with total RSA score (R =  −0.374; p < 0.0001) (Table 3), as well as with the three RSA dimensions: individual dispositions (R =  −0.441; p < 0.0001), family coherence (R =  −0.139; p < 0.05), and social support (R =  −0.186; p < 0.05). Cumulative ACE correlated negatively with total RSA score (R =  −0.309; p < 0.001) and the three RSA dimensions: individual dispositions (R =  −0.258; p < 0.001), family coherence (R =  −0.243; p < 0.001), and social support (R =  −0.291; p < 0.0001). Age was significantly and negatively associated with the RSA social support dimension (p < 0.01), while BMI had no influence on these parameters. There was a significant correlation between the number of current psychological disorders and cumulative ACE (R = 0.20, p = 0.002).

Table 3 Correlation between resilience, age, BMI, cumulative adverse childhood experiences and number of psychological disorders
Full size table

Multivariate analysis

Two models of multivariate analysis were tested to explore parameters significantly and independently associated with total RSA score or with each RSA dimension score. The first model (model A) included the quantitative parameters, namely “number of psychological disorders” and “cumulative ACE”. The second model (model B) included each type of ACE and each current psychological disorder separately (Supplementary Material 1).

In model A (Table 4), the number of psychological disorders accounted for 14% of the variance in total RSA score in the first step. In the second step, cumulative ACE explained a further 6% of the variance in RSA. The final model explained 19.7% of the variance in total RSA score (F = 24.1, p < 0.0001; adjusted R2 = 0.19). Regarding the 3 RSA dimensions, the number of psychological disorders explained 19.5% of the variance in RSA individual disposition score, while cumulative ACE contributed a further 3% of the variance in this score. The final model explained 22.4% of the variance (F = 28.5, p < 0.0001; adjusted R2 = 0.22). The number of ACE was the only parameter to predict the RSA family coherence score (adjusted R2 = 5%). A three-step model explained 13.4% of the variance in RSA social support score. Cumulative ACE contributed to 8% of the variance explained, age to an additional 3%, and number of psychological disorders to 2%. Model B explained 19% of the variance in total RSA score, with sexual abuse, depression, and physical neglect as major factors. The detailed results are presented in Supplementary Material 1.

Table 4 Multivariate analysis (forward selection) predicting total RSA score as well as each RSA dimension (Model A)
Full size table

The mediation model was used in four separate instances, for resilience (RSA total score), individual disposition, family coherence, and social support, to explore whether cumulative ACE (independent variable) affected the number of psychological disorders directly or indirectly. Only significant mediation analyses are presented in Fig. 1. The effect of cumulative ACE on the number of psychological disorders was partially mediated by resilience [RSA total score, indirect effect = 0.107, SE = 0.032, 99% CI (0.038, 0.199)]. With regard to the three RSA dimensions, the relationship between cumulative ACE and the number of psychological disorders was mediated only via the RSA individual disposition score [indirect effect = 0.106, SE = 0.035, 99% CI (0.027, 0.205)].

Fig. 1
figure 1

Mediation analysis model. Relationship between cumulative adverse childhood experiences and number of psychological disorders as mediated by resilience (RSA total score) (upper) and individual disposition (lower). ap < 0.01; bp < 0.001

Full size image

Discussion

To the best of our knowledge, this is the first study in which a significant and independent relationship is demonstrated between a self-report resilience scale for adults (RSA) and both childhood maltreatment and psychiatric comorbidities in patients with severe obesity. In this specific population of patients seeking bariatric surgery, the prevalence rate of ACE (96% patients had ACE), anxiety (36%), BED (35%), substance use (30%), depression (15%), and PTSD (11.5%) was very high, in good agreement with previous studies or meta-analyses [3, 4]. Accordingly, a history of both childhood abuse and lifetime PTSD was also shown to be associated with more severe psychological disorders than PTSD alone in 3045 candidates for bariatric surgery as demonstrated by Walsh et al. [40].

In the present study, among all ACE and current psychological disorders, the total RSA score was particularly negatively influenced by childhood sexual abuse and physical neglect on one hand, and by depression and BED on the other. These findings are consistent with those from studies involving other populations [20, 21, 41]. The RSA dimension of individual dispositions was also negatively linked with all current studied psychopathologies, as well as with history of sexual abuse and physical neglect. These results underline that the presence of current psychopathologies, which is a sign of the psychological vulnerability of patients with severe obesity [42], is concomitant with a decrease in intrapersonal factors of resilience. In addition, these findings confirm the links between ACE and intrapersonal factors of resilience [43]. These resilience factors (family coherence and social support) were only negatively linked with childhood maltreatment, thus emphasising the role of adverse childhood events on interpersonal functioning in adulthood as clearly demonstrated by Poole et al. [44]. Our results are furthermore the first to highlight the link between ACE and psychological disorders with intrapersonal and interpersonal factors of resilience in patients with severe obesity.

The negative relationship between resilience and cumulative ACE is consistent with the dose–response theory whereby an increase in ACE dose has a cumulative effect on mental health and reduces resilience [41, 45]. As in other populations, cumulative childhood maltreatment in patients with severe obesity is negatively linked to intrapersonal [43] and interpersonal factors of resilience [44]. These observations underscore the importance of exploring the presence and number of ACE, and their consequences on intrapsychic and interpersonal functioning of subjects with severe obesity. Furthermore, during the treatment of severe obesity, they emphasise the need for psychological or psychotherapeutic support focused on these ACE and their effects. The negative relationship between resilience and the number of psychological disorders is also consistent with the dose–response theory [46]. Thus, as expected, the number of psychological disorders was negatively correlated with intrapsychic resilience (individual disposition). These disorders were also observed to be associated with a low quality of perceived social support. In the present study, ACE not only had a direct effect on psychological disorders, but also had an indirect effect on psychological disorders through resilience, indicating that resilience was a partial mediator between ACE and psychological disorders. These results are consistent with a previous study which indicated that resilience played a mediating role between ACE and depressive symptoms in Chinese children [47]. Our results also fully agree with those of Spinosa et al. conducted in a general population including 20% of subjects with obesity, in which resilience score was found to be an independent predictor of psychological distress (i.e. anxiety, depression stress) [48]. High levels of resilience might enable patients to take full advantage of personal positive resources [49] and adapt successfully to adversity [50].

The evaluation of resilience by the RSA is consistent with a multidimensional approach of resilience, taking into account its internal dimensions such as individual dispositions, as well as external dimensions including family coherence and social support. Resilience can be considered as an integrative paradigm in regard to its “huge potential for theoretical integration of several fields and issues such as self-regulation and self-stabilisation, coping research, personality development” [51], allowing understanding the emergence of a successful adaptative trajectory after adverse experiences. Thus, taking into consideration that obesity is a chronic multifactorial relapsing disease and given the heterogeneity regarding both the causes and consequences of obesity, a resilience score could prove to be a valid and useful tool to identify clinically relevant psychosocial phenotypes of obesity. High resilience is considered in this instance as a key protective factor for better social, emotional and behavioural functioning [2]. Our patient sample presented a higher resilience score (5.16 ± 0.87) than that observed in other clinical populations, despite a high prevalence of ACE and psychological disorders. Indeed, reported total RSA score was lower in some studies, for example in patients with mood and anxiety disorder (4.11 ± 0.85) [52], in bipolar patients with (3.75 ± 0.7) and without trauma (4.25 ± 0.60) [53], and finally in patients with major depressive disorder (3.04 ± 2.95) [54]. Our results approximate those observed in control populations, e.g. 5.15 ± 0.71 in the study of Hjemdal et al. [29] or 5.15 ± 0.73 in Hilbig et al. [52]. Compared to the population mean, our resilience score appears to be higher than that observed in another study, which used the Connor–Davidson Resilience Scale and showed an intermediate level of resilience lower than that observed in the general population [24]. Several hypotheses may explain these very high resilience scores, especially in patients with psychological disorders. High resilience scores could be associated with positive adaptative coping strategies as recently suggested by two studies [23, 24]. Further studies are needed to explain this process in a population who has faced significant abuse and neglect and high rates of psychological difficulties. These high scores may also indicate post-traumatic growth in certain individuals, i.e., reflecting the perceived benefits in relation to changes in self-perception, interpersonal relationships and life philosophy of the patient after adversity [55]. It is also possible that the socially desirable response in the self-administered RSA could have led to our results, since the RSA is one of the questionnaires used for initial assessment of bariatric surgery candidates. Patients could have considered that RSA scores were used as selection criteria for bariatric surgery, although this was not the case. Furthermore, this can be viewed as defensive functioning or social desirability responding which may potentially invalidate the preoperative mental health evaluation for bariatric surgery candidates [40, 56]. Thus, further studies are needed to analyse the reliability of RSA in patients with severe obesity, by comparing resilience scores with desirability/validity scales (such as MMPI-3), and post-traumatic growth indicators. The use of structured clinical interviews could prove useful, as well as longitudinal data on post-operative outcomes as indicated previously, to improve the understanding of the relationship between resilience and obesity determinants.

From a clinical standpoint, the present results highlight the importance of better identifying ACE and better diagnosing psychological disorders to propose an individualised and tailored psychological support to each patient. The goal is to strengthen protective behavioural factors (such as coping), in order to reduce symptomatology, but also and jointly, to reinforce intrapsychic factors [46], such as mentalisation with two dimensions. First, verbal elaboration of affect is an aspect of mentalisation that refers to the ability to transform affective experiences into communicable representations (symbolisations) shareable with others [57]. Second, reflexive function concerns the capacity to understand other people’s minds as well as his/her own [58]. Mentalisation allows individuals to recover from adverse life experiences without being disorganised [59]. Thus, mentalisation is a process that supports resilience and reduces mental disorders [60]. This dual valence for therapeutic objectives should foster the emergence of resilience dynamics over the long term. The benefit of such support could allow individuals to lessen failures after bariatric surgery, such as weight regain often linked to a resumption of eating disorders, as well as reduce the prevalence of psychological disorders after surgery [61].

Strength and limitations

Our analyses feature several strengths and limitations. Many studies have established the convergent validity as well as the clinical utility of the RSA, especially in differentiating psychiatric and non-clinical profiles [15, 35]. Cross-cultural studies in Brazil and Belgium found that RSA has a very good reliability (Cronbach: 0.88 and 0.84) and has been found to have cross-cultural validity [62, 63]. Moreover, this scale received one of the best psychometric ratings among nineteen reviewed resilience measures, despite the fact the quality of these questionnaires might be considered as only moderate [34]. The second strength of our study is that the evaluation of ACE and psychological disorders was performed by the same experienced psychologist specialising in severe obesity management. Third, patients with severe obesity seeking bariatric surgery have a higher prevalence of psychological distress, in particular depression and binge eating disorder [64], but also of history of ACE [4], compared to other patients who do not seek surgery.

The first limitation of this study is the lack of a control group. It could be argued that findings from patients seeking surgical treatment may not be representative of those with severe obesity seeking a non-surgical approach or even of individuals in the general population, and thus this sample population may overestimate comorbid pathology. However, the purpose of our study was to specifically assess the level of resilience in bariatric surgery candidates with a high lifetime prevalence rate of psychological disorders, including BED and depression. Indeed, patients seeking a non-surgical approach often have a lower BMI, are of older age, and have less psychological disorders than patients seeking surgery [65]. A second limitation of this study is the lack of data gathered on socioeconomic status. It is well recognised that the distribution of obesity follows an inverse social gradient: the lower the socioeconomic status and education level, the higher the proportion of men and particularly of women who are obese [66]. In the Hemmingsson causal conceptual model, psychosocial factors, including resilience, may act as mediators in the association between socioeconomic disadvantage and obesity [2]. Thus, future research should include the socioeconomic component as a covariate in future models. The present cross-sectional study cannot establish a temporal association between resilience, ACE, and current psychological disorder(s), nor understand changes in these constructs over time. Our study shows only an association, but not causality, between a low level of resilience and the presence of ACE and psychological disorders in patients with severe obesity. Moreover, the use of mediation analyses on cross-sectional data being controversial, it is necessary to replicate this mediation with longitudinal data to determine direction of effect between theses variables.

Further studies are needed to analyse the influence of resilience on health behaviours, i.e. whether a high level of resilience, especially after an adverse event, can facilitate changes in health behaviours [67], in particular by reducing eating disorders [20]. Thus, resilience could indirectly contribute to the weight loss of patients with severe obesity during their care pathway. Furthermore, resilience may be a protective factor against the negative impacts of weight stigma, especially on internalised stigmatisation experienced by patients seeking bariatric surgery. Several authors emphasise that weight stigma induces adverse health behaviours and contribute to poor weight-related health, such as maladaptive eating behaviours, low physical activity, physiological stress, and weight gain [68, 69]. Future studies on resilience should incorporate data on internalised weight stigma, to identify whether it mediates weight stigma effects. A longitudinal study may be particularly useful to determine how resilience scores may change with obesity management and to explore some of the reasons why these shifts occur. The present study is the first step of our research program to investigate the change in resilience levels in relation to weight loss in patients treated by bariatric surgery. Due to the dynamic nature of resilience, longitudinal assessments of resilience are of great importance to improve current knowledge regarding this protective psychological mechanism among patients with severe obesity.

Conclusion

The present results highlight an independent negative relationship between both psychological disorders and ACE and the degree of resilience in patients with severe obesity. Our study provides evidence that resilience is a key factor in analysing the psychosocial trajectory of severe obesity candidates for bariatric surgery. A lower degree of resilience is a noteworthy component of psychological phenotypes of severe obesity which encompasses both ACE and current psychiatric disorders (depression, anxiety, substance use, PTSD, binge eating disorder). Thus, the concept of resilience could be considered as an integrative paradigm to better understand the pathophysiology or consequences of severe obesity which is a multifactorial complex disease (involving both genetics on one hand and behavioural, cultural, socioeconomic, and environmental factors on the other). Resilience may play a mediating role between ACE and psychological disorders. Interventional longitudinal studies are needed to demonstrate that resilience is not only a valid prognostic factor for obesity management, but also a critical factor in maintaining mental health after bariatric surgery. The effects of a weight loss management program on resilience scores also need to be established. Further research on the factors promoting the emergence of resilience dynamics, particularly the role of psychotherapeutic guidance offered to patients with severe obesity, is also warranted.

What is already known on this subject?

Although the study of resilience has grown, resilience in the field of severe obesity has not been studied adequately to date.

What our study adds?

To our knowledge, this is the first study examining the link between psychological disorders, adverse childhood experiences and resilience in severe obesity.