INTRODUCTION
The prevalence of severe obesity (defined as body mass index [BMI] ≥ 40.0 kg/m2) is increasing alarmingly in different developed countries in the last years 1,2. However, the reasons for the rapid increase or the possible determinant factors are not well understood. Relatively few works exist addressed to study lifestyle behaviors of severely obese individuals (such as promoting/protecting factors 3)); or consumption of certain food [4,5]), depending on gender. No study was found that characterizes the pattern of food consumption for men and women with morbid obesity prior to bariatric surgery.
Severe obesity entails far more serious health consequences than moderate obesity and is linked with impaired quality of life and health-care spending 6. Nowadays, bariatric surgery procedures are the most successful and durable treatments for severe obesity, as they lead to a significant and sustained weight loss and an improvement in obesity-related comorbid conditions and quality of life 7. Nevertheless, bariatric surgery is not exempt from certain risks 8 and lifestyle behaviors and food pattern identification for planning effective and early preventive interventions for severe obesity remains essential. However, there is little research directly assessing in a globally manner the association between food consumption pattern and severe obesity 4,5,9.
Therefore, the aim of the present study was to describe the food consumption pattern (comparing men and women) of severely obese patient candidates to bariatric surgery and to determine the promoting and protecting factors.
PATIENTS AND METHODS
An observational study was conducted comprising 163 (109 female) severely obese patients who were candidates for bariatric surgery. Their mean age was 45 for both men and women and the BMI was 44.9 ± 6.1 for men and 45.6 ± 5.6 for women (p = 0.06); hypercholesterolemia was found in 31.7% of men and 19.9% of women (p = 0.06); hypertension was found in 85% of men vs 72.4% of women (p = 0.04); diabetes mellitus was found in 28.3% of men vs 19.9% of women (p = 0.18); and smoking habits were present in 25% of men vs 32.1% of women (p = 0.31). Patients were recruited from our reference center and all of them met the bariatric surgery criteria of the National Institutes of Health 10. Subjects provided written informed consent for the procedure and the study protocol was approved by the ethical review board of our institution.
Anthropometric measures such as body weight and height were taken. Individuals wore underwear. BMI was determined as weight divided by height squared (kg/m2).
Food consumption and nutrient intake were measured by a validated food frequency questionnaire (FFQ) 11 administered by a trained interviewer. The FFQ comprised 165 items, including food and beverages. For each food item, participants were asked about their usual consumption of the nine frequency categories, ranging from never or less than once per month to six or more times per day. The FFQ included frequency of consumption of soft drinks (sugar-sweetened carbonated beverages) and fast food products (hamburger, cheeseburger or similar and French fries). A revised Goldberg method 12 was used to identify implausible energy reporters.
The Mediterranean diet score was coded as follows: mMDS: low = 3 (1st tertile mMDS), medium = 2 (2nd tertile mMDS), and high = 1 (3rd tertile mMDS).
Categorical variables were expressed as percentage and 95% confidence interval and continuous variables as mean and standard deviation or the median and interquartile range. Differences in normally and non-normally distributed continuous variables were compared using the Student's t-test and the Mann-Whitney U test, respectively. Categorical variables were compared using the 2 test. Multivariate principal component analysis (PCA) was done to analyze the component that best relates to the food pattern consumption dividing the different food groups in promotors and protectors. For all of the analyses, a p-value < 0.05 was considered as statistically significant, and they were conducted using SPSS for Windows (version 20.0) statistical software package (SPSS Inc., Chicago, Ill., USA).
RESULTS
No significant differences were found between the two groups for age (p = 0.8) and BMI (p = 0.06), whereas height, weight and glycaemia values were higher among men (p < 0.001, p < 0.001 and p = 0.001, respectively).
Lower intakes of fruits (p = 0.006), dairy products (p = 0.087), vegetables (p = 0.003), nuts (p = 0.853), fish (p = 0.04) and Olive oil (p = 0.02) were observed among men, whereas the contrary was found for Spanish food (p = 0.038), soft drinks (p = 0.214) and fast food (p = 0.18) (Table I). Additionally, men reported lower intakes of calories, cholesterol and fiber (p = 0.04, p = 0.03 and p = 0.001, respectively) (Table 1). In the multivariate components analysis, we found that there are two principal components that explain about 70% of variability.
*Data are given either in median and 25th and 75th percentiles or in mean and standard deviation, as appropriate. Units as indicated or in g of nutrient/1,000 kcal. *Men. †Women. ‡Including "paella" (traditional rice dish either with meat, fish, or seafood) and "canelones" (traditional dish of pasta stuffed with meat in a bechamel sauce). §% of energy intake.
The first component that best correlates to the food pattern consumption is the consumption of calories (r = 0.997) and it explains about 40% of variability, whereas the second factor correlates with the diet quality. There is a positive correlation between the food groups that are protective factors for obesity such as the fiber (r = 0.84), vegetables (r = 0.767) and fruit (r = 0.83), and a negative correlation with those factors that are promotors of obesity such as fats (r = -0.341), saturated fats (r = -0.411), soft drinks (r = -0.386), and fast food (r = -0.17). The protective factors are associated to the Mediterranean food consumption pattern whereas the promotor factors of obesity are related to a more occidental food consumption pattern (Fig. 1).
DISCUSSION
The main result of our study indicates that there are significant differences in the food consumption pattern of men and women with severe obesity addressed for bariatric surgery. In other words, the consumption of the different foods and nutrients show significant differences that should be taken into account when planning nutritional intervention. Moreover, a very interesting finding of our study is the relation of the food pattern and the presence of obesity. This finding is in accordance with other cohort studies indicating a favorable and additive effect of lifestyle variables on disease outcomes 13,14. However, and as far as we are aware, this is the first study that has investigated the food pattern differences between men and women with severe obesity.
Regarding quality diet, it is well established that a low adherence to a Mediterranean diet is a risk factor for high prevalence of obesity 15,16. However, studies addressed specifically to a Mediterranean diet and severe obesity are still lacking. In the current study, both groups (men and women) come from a Mediterranean population, but the food consumption related to the Mediterranean diet differs significantly between them. We found that men showed a lower consumption of fruits, dairy products, vegetables and nuts and a higher consumption of fast food and soft drinks than women. This data suggests that severely obese men drift more from a healthy diet pattern than women with a similar BMI.
A previous study 4 had similar outcomes in severe obese patients, concluding that relatively high rates of fast food consumption (three times per week or more) are connected to an increasing rate of severe obesity. Non-specific studies have been done relating consumption of soft drinks and severe obesity, but it is quite well established that an elevated consumption of sugar-sweetened beverages is strongly associated with positively weight changes 17. Specifically, Malik et al. 18 demonstrated that an increase of one-serving per day in sugar-sweetened beverages was associated to an additional weight gain of 0.12-0.22 kg over one year in adults. So, moderate consumption of those food products may exert an important role in preventing obesity.
The current study reinforces the importance of healthy lifestyle behaviors also in obese patients as they play a crucial role in preventing an increase in obesity degree. Moreover, and from a clinical point of view, the identification of such obesogenic lifestyle behaviors and food pattern consumption especially matters in severe obese facing bariatric surgery since the risk of regaining weight after the first two years of surgery depends partially on dietary and lifestyle behaviors 19. However, some limitations of our design and methods should be acknowledged. Food frequency questionnaires measuring past food intake are vulnerable both to random and systematic measurement errors. Although care was taken to provide clear instructions on how to fill out the form, some misreporting cannot be ruled out. Nonetheless, this study has several important strengths, including the high number of severely obese patients participating in the study. Additionally, dietary intake data was recorded using validated questionnaires.
In summary, we have showed that the consumption of different foods and nutrients is different between men and women and should be taken into account when planning nutritional intervention. The current findings might help in the development of public health strategies and are aimed at controlling the alarming increase of severe obesity.