# Childhood Obesity

Background

A foundation of good health established in childhood continues to impact individuals throughout their lives. Unhealthy lifestyle behaviors have been increasingly adopted in the recent times lead to chronic health conditions such as childhood obesity which has proved to be a challenge to American society. Childhood obesity eventually develops into adult obesity which in turn could lead to serious health risks, including stroke and heart failure which in most cases increases health care costs. According to the Dictionary.com, obesity refers to a condition of being overweight or very fat. The USA is currently facing an epidemic of obesity with childhood obesity becoming even more prevalent due to this ongoing issue. It is estimated that there are over 300,000 deaths each year directly related to obesity. Children are more susceptible to diabetes and high blood pressure by their weight gain (Storace, 2013). The main objective of this paper is to identify the factors in the lifestyle of a child that can be controlled in order to prevent childhood obesity in America. The study carried out in this paper uses picked diet, exercise and environment as the most contributing factors in the development of childhood obesity.

**Purpose Statement and Model**

Natural dependent variables of childhood obesity are determined by diet and environment. The most important independent variable in this relationship is diet because in the recent times there has been a shift towards an increased intake of highly fatty and sugary energy foods which at the same time lack minerals, vitamins and other healthy micronutrients. The research will use a theoretical model based on three modifiable variables which include diet, exercise and environment. All the three variables revolve around behavior change which is sensitive to initial condition, non-linear, highly variable, has multiple interactions with and also from both physical and social environments. The model used in this paper involves the use of one dependent variable: diet, the primary independent variable (diet-Pry), exercise and environment. The data used comprises of 30 observations collected from different children within four days. The parameters for determination of obesity are derived from observations in form of short questions mainly on diet-Pry and exercise.. The data is analyzed by first performing a Multiple Linear Regression from which a regression equation is obtained to determine the extent to which each of the variables contributes to childhood obesity.

Definition of the Variables

Childhood obesity refers to “abnormal excessive fat accumulation that presents a risk for health” The same definition is referred to when defining overweight. There is no simple index developed for measuring obesity in children since their bodies undergo several physiological changes in the course of their growth. However, in April 2006, the World Health Organization Child Growth Standards also included measures for obesity for children up to the age of 5 (Global Strategy, n.d). Childhood obesity is caused by multiple factors of complex interaction. When healthy lifestyle and behaviors are not well established at the initial stages of one’s development they become main contributors to the risk of developing obesity. According to the Center for Disease and Control Prevention (CDC) along with National Collaborative on Childhood Obesity Research (NCCOR), state diet-Pry, exercise and environment as the leading factors that contribute to childhood obesity (CDC, 2011).

Diet-Pry is the first independent variable that is presented by many researchers to bear the greatest cause of childhood obesity. Currently, we live in a fast food environment with food we consume being deep fried or frozen. Children tend to eat food without knowing the consequences of their feeding habits. In most cases, a diet-Pry for children implies that a child can consume 40% of his/her daily calorie intake from sugars and fats which lead to obesity. It is stated that obesity costs the US over $136.5billion per year (CDC, 2011).

The second leading factor linked to childhood obesity is exercise. Exercise is understood to be physically active at least 60 minutes every day. The current trend amongst children is passive time spending, for example watching television set. In an effort to promote a healthier lifestyle for children, the first lady Michelle Obama initiated creation of a program called Let’s Move (Move, 2013). It is recommended by the American Heart Association that a child has 60 minutes of daily playing activities to maintain a healthy lifestyle along with proper diet (American Heart Association, 2013).

Environment, which refers to a great range of aspects in nature is the third equally important variable closely related to childhood obesity cases in America. The word environment, in this case, refers a child’s home, child-care centers, community, community and transport infrastructures. The term environment may also refer to the dietary environment in which children have an extensive access to extremely palatable food in large portions. The environment may be measured in terms of time a child to play or watch the television, amount of supervision of school gymnasiums, in the middle schools and the number of programs and recreational facilities in proximity to the children under study (Research Directions in Childhood Obesity, 2013). The first catalyst is the parent’s choice of food purchases is dictated by financial capabilities of parents. Financially challenged parents have a greater tendency to buy more junk foods with no regards to any information of the quality or its consequences to the health of children. Other environment factor is the social settings that a child is in, for instance, if the child’s friends consume sweets on a daily basis, a child becomes more prone to consuming the same products in his diet in order to fit to his environment (Storace, 2013).

**Data Description**

The data has two main sections: dietary behaviors and physical activity. Questions to obtain data on each of these variables are used. The data used in this paper was obtained from a survey titled** “**United States, High School Youth Risk Behavior Survey, 2011” carried out in the United States. The data collected on dietary behavior and physical activity is obtained from the information provided by children and included descriptions and frequencies, such as dietary behaviors seven days before the survey was carried out. There are exceptions where the time duration before the survey is not indicated such as in overweight as well in the physical activity time series data which mostly relied on the general frequency of physical activity. The data obtained was gathered in four days and the total of the average results calculated (See appendix A). One of the limitations of the data presented is lack of the environment variable data. The data presented may not be valid in case the information is obtained from children who cannot provide proper information on diet-Pry over the durations given before the survey as well frequencies. The data is also limited on grounds that it does not include background information on why some children take some diet-Pry.

**Presentation and Interpretation of Results**

Analysis of the data encounters multiple linear regressions since it involves three independent variables which are: diet-Pry, exercise and environment. It is one of many quantitative methods used in prediction of health outcomes in public health surveillance. (Appendix B: The summary Output of the multiple linear regression)

Regression analysis

The general model was regressed as follows:

DEPENDENTVARIABLE : CHILDHOOD OBESITY ADJUSTED R | |||

Independent Variables | Coefficient | T Statistic | Significance of t |

Diet-PRY | -0.19179 | 1.467539 | 2.019258 |

Exercise | 0.218968 | -1.38925 | 0.18861 |

The resulting estimated equation would be:

Childhood Obesity = 0. 218968 Exercise - 0.19179 Diet-Pry

In the situation described above, a high value of R –squared would not have been necessarily relevant and, hence, it would be wrong to think in the course of analysis that if the regression value of R-squared is not more than 80% then it is not right. The other explanation for a small value of R-squared is due to over-use of Excel as a belief that it is the only statistical tool. When entering the data for regression to the scatter plot in Excel, it may give only the R-squared value which is only an indicator of regression model completeness. Only the p-value of coefficients and that in the regression ANOVA table is to be used to determine the value of regression. A value of p-value being less than 5% shows that the regression needs to be considered to have acquired a significant correspondence.

However, it's usually not a good idea, to draw conclusions regarding coefficients when the progression analysis results have given a low value of R2. The reason is that in case there is mis-specified of a model or a wrong variable is added to the model, or a variable is omitted, as in this case, a variable “environment” should have been included, and then the estimator will be biased. The P-stats and Z-values will also be distorted. The same danger lies when the R2 is .09, which implies that the model is lacking some important variables (Haynes, 2010).

The adjusted R^{2 }value, 0.156346, is lower than the Multiple R^{2 }because reflects the complexity of the model and refers to how the number of variances relates to the data. The latter is derived from the regression equation for model performance quantification. It also indicates the variability amount that is accounted for in the data. The value signifies that more than 15 % of the variation is explained by both diet-Pry and exercise. A good choice of a model would contain higher adjusted values of R^{2 }which can pass most or all of the statistical checks. The low value 0.156346 is an implication of many influences on the development of childhood obesity that were not captured in the explanatory variables used in this model. Such influences may include psychological constructs, which are complex to capture in a research model, such as sensitivity and the warmth of parent-child interactions. This probably explains why diet-Pry is always thought to be the primary independent variance (Bigman & Fofack, 2000).

**Discussion of Results**

Linear relationship is either represented as positive or negative. From the regression equation, it can be seen that the explanatory variables have different relationships with the dependent variable since they have different signs. Positive sign on the coefficient associated with exercise indicates appositive relationship between the dependent variable childhood obesity and the explanatory variable, which is exercise. This implies that the behavior of children towards physical activity impacts their development of obesity. On the other hand, the negative sign on the coefficient associated with diet-Pry means that there is no relationship between diet-Pry and childhood obesity. This implies that diet-Pry does not, of its own, contribute to the development of childhood obesity. The coefficients are relatively large indicating a strong positive or negative signs between the two explanatory variables and the independent or explanatory variables, diet-Pry and exercise.

From the empirical results obtained using the empirical model discussed earlier, the value of R^{2 }helps to measure the appropriateness of the explanatory diet-Pry, exercise and the environment modeling childhood obesity in question. From the above results, R^{2 }is 0.309737 which translates to 30 % of childhood obesity. Thus, diet-Pry and exercise indicate 30% contribution to childhood obesity. The R^{2 }may, thus, trigger controversial result since it is commonly known that the two variable are to blame for rising cases of childhood obesity. This result was obtained due to reduction of data related to diet-Pry in order to have an equal number of rows and columns for easy regression. It can, therefore, be significant in forecasting the possible changes that can be applied on the diet-Pry, exercise and the environment of the children to ensure these factors do not become major contributors to childhood obesity. Data analysis allows manipulation of data in various ways with an objective of achieving various predictions.

The R^{2 }is 0.309737 means that 30 % of the variation can be explained by the regression model in the dependent model. The significance of this factor is that the independent variables chosen and the regression model do not support the idea that the variables are not entirely responsible for childhood obesity. In other fields the values of R2are as high as 80 and 90 percent which is a requirement for R2 to be notable. Thus, the judgment of this model is that it was well chosen but the design of the survey was not well structured, which will further be discussed in recommendations part of this paper. The small value of R2 as an outcome should not be used to dismiss the model chosen..

The P-values are computed probabilities for the explanatory variables. The null hypothesis for a statistical test states that a “coefficient is not significantly different from zero”. The P-value for diet-Pry is 0.198165. This value indicates that the probability of the diet-Pry is an effective predictor, hence, it is statistically significant and cannot be removed from the model. It means that there is a 19 % chance that the coefficient of the diet-Pry independent variable emerged randomly and there is no relationship between the two (Motulsky, 2010). Similarly, the P-value of exercise is 0.176286 also indicates a 17% probability of the random relationship. It is also significant and, hence, cannot be removed from the model. There is 81 % probability that the relationship between the independent variables and their coefficients is real. The F-value of 2.019258 used to test the overall significance of regression model and, hence, it presents probability that the model as a whole, emerged randomly. As for both, the lower the significance of P- and F- values, the greater the probability that the relationship between coefficients and independent variables is real and, hence, the relevance of the variable in the model (Regression Analysis, n.d).

**Summary**

The objective of this research was to study the main factors contributing to childhood obesity. The study carried examined thirty observations based on the children behavior with regard to their diet-Pry and physical activity. Indeed, diet-Pry, lack of proper exercise and environment are America’s enemy number one in development of this complication. Childhood obesity not only deprives a child normal development, growth, learning and socialization, but it also poses a risk of emergence of diseases such as heart failure. The government, as well as nongovernmental organizations has identified the three factors as the main contributors to this problem.

From the results obtained, exercise has been empirically identified as the most obvious cause of childhood obesity as can be seen on the regression equation where its coefficient bears a positive sign. Being and independent variable, if a child engages in substantial exercise, then regardless of what the child eats, he or she will not develop obesity; and respectively, lack of favors development of obesity.. Diet-Pry is the primary independent variable which is believed to top the list of the factors responsible for obesity.

However, from the results obtained using a sample of 12 observations instead of 30 initially pre-planned, indicate that there is no direct relationship between diet-Pry and childhood obesity due to the negative sign bone on the coefficient. This tendency is quite disagreeable since it has been proved from other experiments that diet-Pry plays a role in obesity. However, as suggested in (Bigman & Fofack, 2000), this can be true to some extent since lack of exercise and other factors provided by the environment alone can cause obesity regardless of diet-Pry. The limitation in this study was that it was difficult to perform Multiple Linear Regression using MS Excel since the data of variables given could not fit into equal rows and columns.