Traumatic Life Events, Violence, and Obesity: A Cross-Sectional Study from 408 Patients Enrolled in a Bariatric Surgery Program Open Access

In recent years, the prevalence of obesity has been steadily increasing, including in developing countries [1]. Obesity increases the risk of many chronic diseases, decreases quality of life, and increases health care costs, making it a major public health problem [2]. Initially seen in a binary way as a simple energy balance disruption linked to poor lifestyle choices, obesity is now considered to result from a complex genetic and environmental combination, associating to these behavioral factors, several genetic phenotypes, neuroendocrine disturbances, environmental factors (pollution, use of processed foods, etc.), iatrogenic factors (psychotropic drugs, antiepileptic drugs) [3, 4], and epigenetic changes [5].

Recent studies suggest that psychosocial trauma, particularly childhood sexual abuse or violence, is a risk factor for the development of obesity in adulthood [6‒10]. A dose-response relationship between the severity of psychosocial trauma and the risk of obesity has also been reported [11]. Several studies of bariatric surgery candidates have reported a positive association between lifetime traumatic experiences and dysfunctional eating habits, relatively common psychiatric conditions, and postsurgical weight loss in adults with obesity [12‒15]. However, this literature has been limited by small sample size, high variability in the quality of assessment of traumatic experience, and assessment of only certain types of traumatic experiences over the life course [12‒15]. Yet, ignoring the psychological factors involved prevents a holistic patient approach and reduces the treatment’s effectiveness by allowing psychological barriers to persist. The objectives of the present study were to describe the situations of violence and hardships reported by patients undergoing a bariatric surgery program, with a particular focus on sexual violence (SV).

In this cross-sectional study, patients with grade II or III obesity admitted to our digestive surgery department for bariatric surgery from August 01, 2019, to December 31, 2020, either before or within 1 year of the bariatric surgery, underwent a structured interview by a trained psychologist, lasting about 30 min and conducted in 3 consecutive parts. During the first part of the interview, the psychologist explored the family, social, and professional context and invited the patients to describe their feelings about it; this approach allowed the patients to report on the main elements in their lives that generated distress and talk about them with the psychologist. Traumatic life events were recorded within this part of the interview. In a second part, interview focused on weight history, exploring the patients’ ability to describe the circumstances of their weight gain and identifying emotional eating and its links with weight and life history; then the third part was an open interview on possible therapeutic interventions. All interviews were performed by a single interviewer.

As we focus on patient-reported outcomes, the criterion used to identify events as traumatic was the self-identification by the patients of significant suffering. Two types of traumatic life events were analyzed: (1) reported violence, where the traumatic situation was due to intentional use of physical force or power, threatened or actual, by another person or persons; (2) reported hardships, where the traumatic situation was not the result of intentional abuse but of unfortunate life events. Violence was classified as SV (including childhood violence and adult SV and defined as sexual behavior without the person’s consent, including sexual touching, assault, or rape), domestic violence (defined as physical or psychological violence, including deprivation of liberty, psychological control, repeated humiliation, from the intimate partner, but excluding SV), family violence (defined as physical or psychological violence, including deprivation of liberty, psychological control, repeated humiliation, within the family, but excluding SV and intimate partner violence), occupational violence (defined as physical or psychological violence including deprivation of liberty, psychological control, repeated humiliation, at work), and other types of violence (defined as other situations of physical or psychological violence). Hardships were grouped into the following categories: bereavement, psychiatric conditions, medical conditions, sports cessation, work-related distress, burden of a loved one’s hardship, pregnancy, addiction cessation, uprooting, and other hardship events.

Source data were captured in patients’ hard-copy medical records, then entered into an anonymized database for statistical analysis. Patients were classified according to the total number of traumatic life events reported (none, 1, 2, 3, or more).

The primary endpoint was the history of any SV; the secondary endpoints were any hardship and the total number of traumatic life events (violence and hardships). Clinical and demographic data were age, sex, weight, height, socio-professional category according to the INSEE (Institut National de la Statistique et des Etudes Economiques) classification (see online suppl. material; for all online suppl. material, see https://doi.org/10.1159/000535067), the existence of a previous psychological follow-up (distinct from the psychological consultation of eligibility for surgery), and the stage of the surgical process (pre- or postoperative).

Mean (standard deviation) and number (frequency) were used to summarize the data. To test the association between socio-professional category and history of psychological follow-up across gender and other between-group comparisons, Welch’s t tests to account for unequal variance between groups and Fisher’s exact tests were used. The association between BMI and the total number of traumatic life events was performed using linear regression. Multivariate logistic regressions were applied to identify independent risk factors for SV and any hardship using R (4.0.3) software. To take account of multiple comparisons, we have used the Bonferroni correction by dividing the alpha error by 3, with a significant level of 0.017. The study protocol was approved by the University Hospital IRB of Montpellier (IRB-MTP_2021_09_202100928), and consent was obtained from all participants.

Data from 408 patients with obesity (73% women, mean age 45.9 years, mean BMI 40.5 kg/m²) were analyzed (online suppl. Fig. S1). For participants in presurgery course, the median time between interviews and surgery was 1 year (IQR, 0–1), with a range of 0–4 years, while for those in post-surgery course, the median time between interviews and surgery was 3 years (IQR, 2–6), with a range of 1–17 years. Nearly one third of them were or had been followed by a psychologist (Table 1), and this was independently associated with sex and socio-professional category (online suppl. Table S1). Regarding violence, the most frequently reported were sexual, family, and domestic violence, while bereavement, psychiatric conditions, and health status were the most frequently reported hardships. Compared to men, women reported experiencing three times more violence in all categories (p < 0.001), including four times more SV (p < 0.001) and seven times more domestic violence (p = 0.001).

Women who reported SV had a higher BMI (41.5, 7.8 vs. 38.9, 7.1; p = 0.003). Figure 1 shows that regardless of age, socio-professional category, and psychological follow-up, gender and BMI were significantly associated with an increased risk of SV history. Multivariate logistic regression indicates that this risk is multiplied by 7.4 for women (p < 0.001) and increased by 5% for any increase in BMI of 1 kg/m² (p = 0.002).

Conversely, women reported fewer hardships than men in all categories combined (p < 0.001). The most common hardships reported by men were work-related distress (p = 0.011), cessation of sport (p < 0.001), and psychiatric conditions (p = 0.039) (online suppl. Tables 1 and S2). Within psychiatric conditions, men reported depression more frequently than women (18.2% vs. 8.05%, p = 0.006), whereas no significant difference was found for eating disorders and mental retardation.

The total number of traumatic life events was similar between men and women (p = 0.266). Overall, linear regression indicated that BMI increased by 1.38 ± 0.47 kg/m² for each increase in the number of traumatic life events (p = 0.0036) and compared with patients reporting one or more events, BMI increased by 2.19 ± 0.86 kg/m² (p = 0.0118) and 3.06 ± 1.29 kg/m² (p = 0.0178) for patients reporting two and three or more events, respectively (online suppl. Fig. S2). Multivariate logistic regression indicated that the risk of having experienced at least one hardship was increased by 3.2-fold (adjusted odds ratio OR = 3.22, 95% confidence interval: 1.85–5.88; p < 0.001) for men compared with women, independently of age and BMI (online suppl. Fig. S3).

Our results show that, among patients included in a bariatric surgery program, women report a higher frequency of SV, whereas men report a higher frequency of hardships related to psychiatric conditions, bereavement, working conditions, or life events in general. The proportions of traumatic life events reported in our study are very high but comparable to those reported in other patients with nutritional disorders. In a meta-analysis involving a total of 13,059 patients with eating disorders, childhood maltreatment prevalence rate ranged from 21 to 59%, relative to 1–35% in 15,902 healthy subjects [16]. Fewer data are available in bariatric surgery patients. However, a recent systematic review and meta-analysis reported a prevalence estimate for at least one form of adverse childhood experience of 51% (95% confidence interval: 32–70%) among a total of 1,368 patients [ref: PMID 36270937]. Our results confirm the high prevalence of lifetime adverse experiences in bariatric surgery patients.

In particular, one third of our patients declared having been submitted to SV, a rate higher than that reported in the VIRAGE (violence and gender relations) population-based survey, conducted in 2017 among 15,556 women and 11,712 men, aged 20–69 years and living in ordinary households in metropolitan France. In this survey, one in seven women (14.5%), and one in 25 men (3.9%) declared having experienced at least one form of sexual assault in their lifetime, while reported SV is 2.9 times and 2.6 times higher for women and men, respectively, in our population. This difference is likely due to the fact that our population includes only individuals with obesity, as we also show that there is an independent association between BMI and SV.

The cross-sectional nature of our study precludes determining the direction of the possible causal relationship. Several prospective studies have shown that childhood abuse may have an impact on subsequent overweight [6, 7]. Fleischer and colleagues have shown that physical and emotional neglect and childhood sexual abuse are associated with a higher risk of overweight in both sexes, with a stronger impact of physical neglect in males and a stronger impact of emotional neglect in females [6]. In another survey of 2,211 participants, the proportion who reported childhood abuse was 24%, 26%, 29%, and 37% among those classified as overweight and with obesity class I/II, and III, respectively [7]. In addition, these abused participants reported that their weight problems started earlier, were more likely to have weight-related comorbidities, felt judged by health professionals, were less likely to feel treated with respect, and had lower self-esteem [7]. The association between abuse and weight gain can be direct, as the acute or chronic stress may trigger a cascade of physiological responses activating the cardiovascular, metabolic, and inflammatory systems [17‒19], or indirect, as the trauma can induce various behavioral changes (unbalanced diet, eating disorders, reduced physical activity, addictive behaviors) [20, 21], all related to the risk of obesity and diabetes [14‒16]. However, obesity itself could induce psychological fragility, making the person more at risk of experiencing situations of abuse.

Our study also highlights the importance of reported hardships among the men involved in a bariatric surgery program, as almost 85% of them reported at least one situation of hardship, men being three times more likely to report at least one hardship than women of the same age and BMI. As sport cessation is the most frequently mentioned, we suggest this to be considered in their management with more exercise rehabilitation or adapted physical activity programs. Surprisingly, the other hardship situations more frequently reported by men were work-related distress and depression [22], two situations generally reported more frequently by women than by men in standard populations. Although the cross-sectional design of our study precludes the possibility of assessing causality, these results suggest that psychological distress might have specificities in the context of obesity.

Despite the difficulties encountered, only 32% of our patients reported having received psychological follow-up. Our results show that, independently of the trauma, the use of such follow-up depends on gender and socio-professional category. Understanding this reluctance would make it possible to develop other types of care adapted to these situations and to make patients with obesity aware of the possibility of a link between their current state of health and a possible history of abuse suffered in childhood or at any stage in life. Training of professionals working with this population is therefore essential.

The strengths of our study include the comprehensive assessment of lifetime traumatic event history in a large sample of men and women by a trained and experienced professional. In addition, the nonparticipation in our study of 8% was low, minimizing selection bias and increasing the representativeness of our sample to the entire target population. Nevertheless, several limitations may be acknowledged. The data used in our study comes from a single center, which limits the generalizability of the results to all individuals with obesity who are candidates for bariatric surgery. These are patient-reported outcomes, so we cannot verify the reality of the events, nor can we eliminate a recall bias. The cross-sectional nature of the study prevents the assessment of a temporal link between the events reported and obesity.

Recognizing the impact of traumatic life events reported by patients enrolled in a bariatric surgery program is a critical issue in providing them with the best management options. Raising awareness about the extent of the psychological burden suffered by these patients might help to reduce the negative perceptions and stigma associated with obesity.

We thank Perrine Mosset for helpful data collection.

The study followed the precepts of Ethics in Research with Human from the Declaration of Helsinki and was approved by the Institutional Review Board (IRB: IRB-MTP_2021_09_202100928) from the Research Ethics Committee of Montpellier. Since the study used patient-level data and directly interviewed subject, the written informed consent was obtained from all of them.

All authors declared no competing interest.

This study was not funded.

A.D.: study design, data analysis, and writing; N.M. and D.N.: study design and data collection; J.B.B.: study design; C.B., V.L., A.R., P.L., S.J., A.A., and E.R.: data collection, and F.G.: study design, data collection, and writing. All authors discussed the results and approved the final version of the manuscript.

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