Prevalence and Factors Associated with Overweight and Obesity among Adolescents in Schools in Benin in 2016

Introduction Overweight and obesity are considered to be global public health concerns. The objective was to estimate the prevalence of overweight and obesity and the associated factors. Methods This was a cross-sectional study conducted in 2016 and involving 3968 adolescents of 10 to 19 years from 40 secondary schools in Benin. Participants were selected using a 2-level random sampling technique. Data were collected through a self-administered questionnaire and were analyzed using Epi Info and Stata softwares. Associated factors were investigated by a logistic regression. Results Mean age was 16.54 ± 0.04 years and sex ratio was 1.16. The prevalence of overweight was 8.1% and that of obesity was 1.6%. Associated risk factors were female gender, sedentary lifestyle, consumption of sugary and carbonated beverages. Consumption of fruits and vegetables, regular physical activity, age over than 15 years were protective factors. Conclusion Overweight and obesity among school adolescents are increasing in Benin. The establishment of a school nutrition program and the promotion of regular physical activity should contribute to reducing their prevalence.


Introduction
The stable economic development of a country depends on the level of education, the state of health and the optimal well-being of adolescents and young people.
Childhood and adolescence are recognized as periods of learning about health practices and behaviors, particularly in nutrition. Adolescence is the most important period of health. It is undermined by many health problems including overweight and obesity. The World Health Organization (WHO) considers adolescence to be the period of growth and human development that is between childhood and adulthood, between 10 and 19 years of age. Overweight and obesity are serious public health concerns to adolescents.
In fact, overweight and obesity in an adolescent, later predispose him to chronic diseases such as cardiovascular diseases, diabetes, chronic respiratory diseases and cancers [1]. Other consequences of obesity are psychological effects and social stigma faced by obese adolescents, which can have serious consequences for mental and physical health [2].
Some factual data shows that the school environment influences the consumption of foods and is likely to promote unhealthy eating habits that promote the onset of obesity among students [3]. Besides eating habits, these statuses are as well conditioned by factors such as the practice of physical activity and the involvement of parents, as reported in some studies in developed countries [4].
In addition, the prevalence of overweight and obesity among adolescents has increased steadily in the different age groups and in most countries in the past decades [5]. Thus, worldwide, in 2010, about 42 million children were overweight with 35 million living in developing countries [6]. The prevalence of overweight and obesity among children and youth is increasing in emerging countries in Africa and Asia [7].
Overweight and obesity in adolescents, formerly associated with wealth and developed countries [8] [9], also affect developing countries. In sub-Saharan Africa, among school-aged adolescents, the prevalence of overweight and obesity varies from one country to another. For example, in Nigeria in 2012, 9.7% of adolescents were overweight and [10] while in Senegal in 2011, 9.7% were obese [11].
Although several studies are being conducted in Western countries on the subject, this is not the case in Africa where the focus is mainly on communicable diseases such as malaria and HIV/AIDS.
In Benin, the nationwide survey of school-based student health, conducted in 2009 in the framework of the Global School-based Student Health Survey (GSHS) showed a prevalence of 8.8% for overweight and 0.4% for obesity [12]. But since then, no further studies have been conducted in the area to provide updated information for better targeted interventions.
It is therefore to contribute to filling this gap in strategic information that the present study was conducted with the objectives of estimating the prevalence of overweight and obesity and identifying the associated factors among school adolescents in Benin.

Description of the Survey and the Study Population
We analyzed data from the national survey on the current status and factors associated with adolescent health in schools in Benin in 2016. This was a cross-sectional study targeting 10 to 19-year-olds children of both sex and enrolled from first to seventh (final) grades of secondary school. The number of subjects required was calculated by the Schwartz formula and was 3841. The sampling was conducted with the technical support of CDC Atlanta, using a 2-degree random sampling technique with a probability proportional to the size of the schools and classes.
The first level consisted in selecting 40 out of 1494 public and private secondary schools, with a probability proportional to the size of the institutions. At the second level, within each selected institution, a census of all classes was conducted. The choice of classes to be investigated was randomly made from a list of random numbers, pre-established and made available by CDC Atlanta and WHO. This list of numbers varied from a school to another. The number of classes to be surveyed per institution was proportional to the total number of classes available in these selected institutions. Finally, all adolescents of eligible classes present during the survey were included in the study.
The data collection tool was an adapted WHO questionnaire that included demographic and foods behavior data (Annex 1). The weight of each teenager was measured to the nearest 0.1 kg, using an electronic scale (model 753 E, Seca, Hamburg, Germany). The height was measured to the nearest 0.1 cm, in a standing position with a SECA height measure device.

Dependent Variable
The dependent variable was the Body Mass Index (BMI) defined using height and weight. The BMI of students was ranked using the WHO growth reference for school-aged children and adolescents by age and sex. Thus, the Z scores for BMI for age were generated, using the WHO "Anthroplus" software [13].
In accordance with the WHO classification, the BMI was categorized in four modalities: adolescents who had a BMI less than −2SD (<−2SD) for age and sex were considered to be thin and were classified as underweight; those with a BMI between −2SD and 1SD (−2SD ≤ BMI ≤ 1SD) were classified as normal; adolescents who had a BMI between 1SD and 2SD (1SD < BMI ≤ 2SD) were classified as overweight while those with a BMI greater than 2SD (BMI > 2SD) were classified as obese. But for the sake of convenience in the statistical analysis, the dependent variable was further categorized into three main modalities by grouping together the normal and underweight modalities which constituted the reference modality ("normal") in the polytomic regression.

Independent Variables
The independent variables were age, gender, areas of residence, class attended, type of school (public or private), regular pratice of physical activity (exercise for at least 60 minutes a day and at least 3 days a week), consumption of at least five servings of fruits and vegetables, sedentary lifestyle (staying in a sitting position for more than four hours a day), consumption of sugary and carbonated beverages.

Data Analysis
The data were entered with the Excel software and analyzed using the STATA 13.0 software. The qualitative variables were estimated in proportion with their 95% confidence intervals (95% CI) and the quantitative ones in means followed by ±standard deviation (SD).
Associations between qualitative variables were studied using the Pearson Chi2 test. The strength of the associations between the dependent variable ("BMI") and the independent variables was assessed by calculating Odds Ratios (OR) followed by their 95% CI. The "overweight" and "obese" modalities of the dependent variable "BMI" were compared to the "normal" (reference) modality.
A polytomous logistic regression was used to evaluate the influence of the explanatory variables on the BMI. All statistically significant variables at a p-value < 0.20 in the univariate analysis were included in the multivariate models and a stepwise modeling procedure was adopted to select the variables that were used to build the final model. Only variables with a p-value of less than 0.05 have been retained in the final model. To take into account the effect of the sampling procedure used, we weighted all our analyses. The adequacy of the model was verified using the Hosmer-Lemshow test.

Ethical Consideration
The protocol was approved by the Institutional Research Ethics Committee of the Institute of Applied Biological Sciences (CER-ISBA). Legally major students (18 years or older) provided their informed consent while the minor ones (less than 18 years) provided their informed assent along with their parents informed consent. All data were anonymous.

Sociodemographic, Behavioral and Psychological Characteristics of Participants
Our analyses included 3968 school adolescents, which corresponds to a response rate of 78.00% of the sampled population. Table 1 shows the distribution of adolescents according to general characteristics. The average age of the 3968 adolescents included in the study was 16.54 ± 0.04 years and the sex ratio male/female of 1.16. They lived mainly in urban areas (60.5%) and 68.9% were in public schools. More than one in two teens had lived with their parents in the last 30 days prior to the survey ( Figure 1).
The proportions of adolescents who consumed fast food or sugary and carbonated beverages were 38.6% and 34.9%, respectively while 56.4% had regular practice of physical activity ( Table 1).
The distribution of adolescents according to the psychological variables shows that 14.3% of the respondents had often felt lonely in the last twelve months preceding the survey; 20.8% were often anxious during the same period and 15.9% thought of a suicide plan (Table 1).
Some authors have found prevalence of overweight and obesity of the same order of magnitude as ours. For example, Musa & al. found an overweight prevalence of 9.7% in Nigeria among school-aged adolescents in 2012 [10]. In 2009, the national school health survey (GSHS) in Benin reported a prevalence of 8.8% for overweight [12]. In Togo in 2010, Djadou & al, reported a prevalence of obesity of 1.7% [14]. In Ghana, in 2014, Manyanga & al. found a prevalence of 8.7% for overweight and of 1.0% for obesity [15]. The magnitude of overweight in this study may be explained by the nutritional transition taking place in the context of empowerment, which leads to a double nutritional burden [15]. Traditional food tends to disappear in favor of a Western diet consisting mainly of fast food dishes, associated with the consumption of sugary and carbonated beverages and sweet foods. The high prevalence of overweight may also be explained by the fact that in African culture, being overweight is synonymous with wealth and health [15].
Other authors, on the other hand, found higher school prevalence than ours. These include Mantey [19] in, 5.0% [20], 5.8% [21], 9.3% [11], 4.4% [22] respectively, they were all superior to ours. Open Journal of Epidemiology Contrarywise, some authors like Regaieg & al [22] and Benyaich & al [23] found in schools, lower prevalence of overweight than ours, which were of 6.3% and 2.0%, respectively. Obesity prevalence lower than ours was also found in the GSHS studies in Benin (0.4%) [12] and in Malawi (0.8%) [15] both in 2009. The observed differences can be explained by the difference in the target populations and sites. Indeed, none of these studies focused on adolescents aged from 10 to 19 like ours and some of them were conducted at the local and regional levels while others were at the national level. The differences could also be explained by the difference in data collection techniques.

Sex
The female predominance observed in school settings among adolescents for overweight in our study was also reported by several other authors in Nigeria,  [22]. This gender-based difference could be explained, first by the age range of the studied children since, after a pre-pubertal rebound, there is an increase of the fat mass in the girl while that of the boys decreases [22]. At the age of puberty, girls are more likely to develop fat masses related to growth. In fact, adolescent girls have a hormonal predisposition to developing gynoid obesity and are also more sedentary than boys. It could also be explained by the fact that in African households, manual work requiring physical efforts are more frequently assigned to boys than to girls [15].

Age
Our results showed that adolescents aged 15 and older were less likely to be overweight and obese than those under 15 years of age. This association has also been reported in Senegal and Nigeria [10] [11]. Manyanga & al [15] reported a higher risk of obesity or overweight among adolescents under 15 in seven African countries, namely Benin, Djibouti, Egypt, Ghana, Mauritania, Cameroon, Malawi and Morocco [15].
This association with the under-15 age group was explained by the fact that with age, adolescents become more and more sensitive to the mockery of their peers with regard to overweight or obesity. They become more concerned about their body image [25]. They would therefore tend to impose dietary restrictions and/or the practice of physical activity to themselves. By contrast, Manyanga et al in Malawi [15] and the Alsace regional observatory [15] have found an increase in overweight with age.

Regular Practice of Physical Activity
In our results, there was a significant association between overweight/obesity and regular physical activity. Indeed, this variable has been shown to be a protective factor for overweight/obesity. The same observation has been made in most studies involving adolescents in schools, including those of Bhuiyan & al. in Bangladesh [24]; Dendana & al. in 2016 in Tunisia [26]; Badr W & al. in Tunisia in 2016 [27]; Regaieg & al. in 2015 in Tunisia [22] and Desalew & al. in Ethiopia in 2017 [21]. This relationship is biologically expected as physical activities favor the consumption of body fat mass. However, a few numbers of authors like Allam & al. in Algeria in 2016, did not find this association in their studies [6].

Daily Consumption of at Least Five Servings of Fruits and
Vegetables According to our results, consuming at least 5 servings of fruits and vegetables a day was a protective factor for overweight/obesity among teens. Dendana & al. in 2016 found similar results [26]. This could be explained by the fact that vegetables and fruits consumption makes us so rapidly fed up that we do not eat so much.
On the other hand, other authors have found an increase in overweight with the consumption of fruits and vegetables. This is the case of Badr & al. in Tunisia in 2016 among adolescents of the general population [27].

Consumption of Sugary and Carbonated Beverages
In our study, the consumption of soft drinks was a risk factor for overweight and obesity. This same association was found by Desalew & al. in Ethiopia in 2017 [21], St-Onge & al. for overweight [28] and the health monitoring institute in Haute-Savoie [16]. On the other hand, other authors found results contrary to ours. This is Li M & al [29] for overweigh. This difference could be explained by the difference in study settings.

Sedentary Lifestyle
In our study, there was a significant association between overweight/obesity and sedentary lifestyle. In the literature, several authors have made the same observation in school adolescents. These include Regaieg & al. among school adolescents aged 15 to 18 [22]; Bhuiyan & al. among adolescents aged 10 to 15 in Bangladesh [24]; Desalew & al. in Ethiopia in 2017 [21]; the Institute of Public Health Surveillance in Haute-Savoie in France [16]. This could be explained by the modernization of society and the change of social behavior [18], particularly with the exposure of adolescents to new technology [15]. Indeed, teenagers spend a lot of time in front of television and video games [22]. Obesity results from an imbalance between insufficient physical activity and a diet that is too high in calories. Any movement consumes energy and protects against obesity. That is why it is important to move. In addition, these sedentary periods encourage the consumption of important caloric foods (chocolate, chips, peanuts).
On the other hand Dendana & al. in Tunisia in 2016 [26] did not find an association between the sedentariness and the overweight.
Finally, in our study, we were unable to demonstrate a significant association between fast food consumption, hunger, place of residence, the type of school attended (private or public), loneliness, anxiety, trauma, suicide planning, relationship with parents/guardians (benefitting from the parents or guardian's kind attention and listening), having no close friends, the person with whom the teenager lived and overweight.
The differences between the studies cited above and ours regarding associated factors could be partly explained by a lack of power as most of them were of relatively small size. In addition, they were conducted on different age groups and none covered the whole adolescent age range (10 to 19 years), as was the case in our study, and most have been carried out at the local or regional level, making difficult to compare with ours (Table 3(a) & Table 3(b)).

Limits
For logistical reasons, the study could not be extended to university; this could have led to a selection bias, since adolescents are also at university. However, we are certain that most of teenagers are in secondary schools and very few in universities, which could minimize this selection bias.
Family and genetic factors play a role in the development of some obesities but were not taken into account in the present study.
The self-administration of the questionnaire might have induced some information biases due to interviewees negligence or mis-understanding of certain questions. However, we addressed this situation by explaining each question to teens before questionnaire filling. Also, we may have some desirability bias as

Conclusion
Our study shows that overweight and obesity among school adolescents are a serious problem in developing countries, particularly in sub-Saharan Africa including Benin. This overweight is significantly related to factors such as gender, age, regular physical activity, consumption of sugary and carbonated beverages, sedentary lifestyle and fruits and vegetables consumption. Considering that the socio-economic development of developing countries depends upon the physical, social and emotional well-being of adolescents, these results argue for taking individual and collective measures to prevent overweight and obesity among school adolescents and thereby the subsequent onset of chronic diseases in their adulthood. Open Journal of Epidemiology

BENIN SURVEY ON THE STATE OF THE HEALTH FACILITIES OF ADOLESCENTS IN SCHOOL ENVIRONMENTS
This survey is about your health and the things you do that may affect your health. Students like you all over your country are doing this survey. Students in many other countries around the world also are doing this survey. The information you give will be used to develop better health programs for young people like yourself.
DO NOT write your name on this survey or the answer sheet. The answers you give will be kept private. No one will know how you answer. Answer the questions based on what you really know or do. There are no right or wrong answers.
Completing the survey is voluntary. Your grade or mark in this class will not be affected whether or not you answer the questions. If you do not want to answer a question, just leave it blank.
Make sure to read every question. Check the boxes on your answer sheet that match your answer. When you are done, do what the person who is giving you the survey says to do.
Thanks very much for your help.