Eating Behavior and Weight Trajectory After Bariatric Surgery

The objective of this study is to evaluate the association among eating patterns (specifically maladaptive behaviors), certain psychological variables and weight trajectory in the short- and long-term after bariatric surgery.

Design: Five-year, prospective, multicenter study.

The study will include 2 groups:

1. Candidates to primary BS undergoing laparoscopic gastric bypass (LGBP) or laparoscopic sleeve gastrectomy (LSG) from September 2020 to September 2021.The subjects of this group will be followed prior to BS, at 4 months, 1 year, 3 years and 5 years after BS.
2. A control group of subjects with obesity not candidates to BS matched with the intervention group for age, sex and BMI prior to BS. They will be evaluated once.

Inclusion criteria:

• Candidates to primary BS from 09/2020 to 09/2021.
• Subjects with obesity (>40 kg/m2) not candidates to BS.
• Internet users.
• Primary BS: LGBP and LSG.
• Accept to participate in the study and provide signed informed consent.

Exclusion criteria:

• Second time and/or revisional BS.
• Pregnancy.
• Presence of intellectual impairment impeding the administration of the psychometric tools.
• Type 1 diabetes.
• Dyslexia. Weight variables: definitions.

1. The different weight variables will be calculated as follows:

• Weight lost (kg): Pre-BS weight - weight at nadir.
• Current weight lost: Pre-BS weight - current weight.
• Ideal weight: weight corresponding to a BMI = 25 kg/m2.
• Excess weight lost (%): (Pre-BS weight - current weight)/excess weight [(pre- BS weight - ideal weight) x 100].
• Total weight lost (%): (pre-Bs weight - current weight/pre-BS weight) x 100.

1. To study WR the following variables will be calculated:

• Total WR (%): (100/(pre-BS weight - weight at nadir)) * weight regained (kg).
• Excess WR (EWR %): [(current weight - weight at nadir)/excess weight] x 100
• Weight regained (Kg): current weight - weight at nadir.

Excess weight lost > 50%, total weight lost > 25% and EWR > 15% will be considered as significant.

Information regarding the psychological and behavioral variables will be collected using questionnaires that have been validated in our setting and will be completed by the patients themselves online.

To assess hedonic response to the foods, the following questionnaires will be used:

• The subscale of external signal-triggered eating of the Dutch Eating Behavior Questionnaire (DEBQ) that includes 10 questions.
• A food addiction questionnaire (Yale Food Addiction Scale II) with 25 questions.
• The Bulimia Investigatory Test Edinburgh (BITE) made up of 2 subscales of symptoms and severity including 33 questions.
• There will also be questions related to snacking/nibbling.

The following will be used to assess mood state, stress and negative emotions:

• Dutch Eating Behavior Questionnaire (DEBQ) which evaluates eating styles and takes into account 3 subscales: external, restrained (10), and emotional eating (13) including a total of 23 questions.
• Positive and Negative Affect Schedule (PANAS) to assess the positive and negative effects of emotional experiences including 20 questions.
• Barrat Impulsivity Scale (BIS-11) which evaluates impulsive tendencies in 3 dimensions: cognitive, motor and lack of planning and includes 30 questions.
• The consumption of toxic substances will be determined using the following questionnaires: Alcohol Use Disorders Identification Test (AUDIT-4) including 4 questions in which data on the frequency and quantity of alcohol consumed will be obtained, and the Fagerström Nicotine Dependence Scale including 6 questions.
• The Eysenck Personality Questionnaire - Revised (EPQ-R) with only the dimension of neuroticism including 23 questions.

These variables will be considered as confounding psychological variables and will, therefore, be controlled by statistical tests.

• Depression: evaluated with the Hospital Anxiety and Depression Scale (HADS) including 14 questions.
• Perceived economic status, assessed using a visual analog scale of 10 cms with markings at 0, 5, 10 cm without numbers.
• The International Physical Activity Questionnaire (IPAQ) including 7 questions to assess physical activity.
• Social support evaluated by the evaluation protocol using the English Longitudinal Study of Aging study. APGAR (including 5 questions).
• Quality of life evaluated using the Impact of Weight on the Quality of Life - Lite (IWQOL-lite) questionnaire including 31 questions.

Evaluation of food intake: The patients will also complete a registry of foods and drinks consumed over 3 days including a weekend day. It should include the type of food, cooking method, and portion size or weight. Afterwards, the total intake and by macronutrients will be analyzed with the Diet Source software (Nestle Health Science. V4.0).

Primary variable:

Body weight: total weight lost (%), excess weight lost (%), total weight regained (%), excess of weight regain (%).

Data collection:

Additional information on demographic and anthropometric characteristics and on comorbidities both prior to BS and their evolution over time will be collected in an Excel datasheet designed with this objective.

Sample size The study will include 445 patients undergoing BS. It is estimated that the percentage of losses to follow-up at 5 years will not be greater than 10% and thus, at least 400 patients must achieve complete follow-up. This sample size will have a statistical power of 80% for detecting an effect size, that is, differences divided by the standard deviation of at least 0.281. Additionally, 100 controls not undergoing BS will be included and matched for age, gender and BMI with the objective of standardizing the values in the scales. The ratio will be 1:4 for controls and patients undergoing BS, respectively.

Statistical analysis. The design and statistical analysis of the present study meet the recommendations of consensus documents in the literature, especially the TRIPOD statement. A statistical analysis plan (SAP) will be elaborated prior to finalization of data collection and will provide an in-depth description of the statistical methods to be used, the tables and figures that will be included in the statistical report as well as the strategy to follow in the case of missing values, and multiplicity adapted to regulatory and scientific recommendations.