STATA USER ONLY

Assignment 1

Introduction

In this study we will choose Body Mass Index (kg/m2) as the variable to be studied. In the U.S. Department of Health and Human Services profile, Body Mass Index (BMI)1 is associated with heart disease, diabetes, hypertension and other chronic diseases that may cause death. Therefore, focusing on the annual mean of BMI across ages and genders can be effective in analyzing trends and making predictions.

According to the Centers for Disease Control and Prevention (CDC)2 data from the National Health and Nutrition Examination Survey (NHANES) from 2017 to 2018, this data did not differ significantly between adult men and women of the same age, obesity with an age-adjusted prevalence of 42.4% among adults. At the same time, the report suggests that the prevalence of obesity and severe obesity is higher in non-Hispanic races. With regard to possible causes of disease, They showed a low correlation (R2 = 0.067) between obesity rates and myocardial infarction in five states, based on a 2011 article3. The strongest association was found between hypertension and obesity rates (R2 = 0.811), followed by the correlation between obesity and hypertension rates (R2 = 0.7622).

Since the prevalence of obesity is increasing rapidly in children and adults worldwide, the values are generally higher in African-Americans by comparison. A high BMI is easily recognized within the scientific context as having a higher probability of obesity. Therefore, we can propose this hypothesis: African Americans have a higher chance of having an excess BMI on average in the same age group of different men and women.

Furthermore, we can conduct a questionnaire survey of demographics by categorizing food groups by the length of exposure and the presence or absence of healthy content of the total amount of food eaten per day at a given time. This method4 was investigated to determine the dietary patterns of overweight and obese African American women in the southern United States through 30 food groups. Potential confounding factors under this type of approach include, but are not limited to: the inability of families to afford to purchase fresh, healthy foods over time so choosing quick and easy fast foods, geographic factors that make it more difficult to purchase fruits and vegetables and to stay athletic, whether food tastes developed from childhood are high in sugar and salt, and the ability of the questionnaire participant to cook (disability, adolescents). For these effects, we need to take steps to adjust the data5. In this content depending on the country region, race, income, and birth order, this report adjusts the estimates based on income, in different cases according to the proportion of distinct effects.

Method

We will use Stata 17.0 for data analysis. The data is scattered and organized into seven panels of 602 units, categorized by census science according to different directories. However, since the data is not detailed, we need to recreate a new category for generating. At the first moment of getting the data use summarize() to give an overview of the data. Then we had to think about the age of the survey BMI, the part of the African-American population included. We include Non-Hispanic Black in the ridreth1 variable as the only source of African American ethnicity in this report. After these refined process decompositions, we first retain the choice of ridreth1 == 4, and with summarize bmxbmi we obtain a rough generalization of BMI. The mean of the new variables in these 126 observations is 29.98413, which is in the overweight group in the BMI content. By learning about this average, we can find out if there is a difference in the definition by comparing the values of other races. For example, for non-Hispanic Caucasians, although they generally have a greater average age in greater numbers, they have an average BMI of 29.4716. For African-Americans who are younger than the values, they are in better physical health instead.

In this database, the main exposure of interest, the main outcome of interest and all relevant covariates, we can only insinuate the impact of economic status through the distribution of household income to poverty ratio. Following a similar approach we generate an analysis for African Americans: selecting the African American only option and then doing a study on their evaluated income poverty ratio. We found that this figure was 2.18, which is not considered moderate. However, there are far more potential confounders or modifiers to the effect measure than this. We can only see their income numbers, but there are no indicators in the data that capture the type of food consumed, the income distribution of each household with respect to food, or the convenience of their geographic location. Lacking detailed information, we can only venture a conjecture.

When we investigate the expected plan with other methods: in the bi-variable analysis, the results created by the regression model according to different races and BMI: we can see that the race-related probability of BMI (0.004) is much less than 5%. We can review at this point with corr, according to the correlation model BMI and race from rank 1-5, is -0.1183. Thus in this correlation between the two, the analysis of bi-variable confirms that there is a connection in the importance of the data. Nevertheless, when we use multivariable linear regression, the situation is not quite the same. When the outcome of regress is race, and the relation to that is the income poverty rate, we see that the R2 of the whole model is 0.006, while the probability is 5.2%. This proves to be just stuck at the 5% line rubbing off. When my exposure is race and outcome is BMI, despite the small R2 number, there is no denying the strong significance of these two in the data.

References

1. Assessing your weight and health risk. National Heart Lung and Blood Institute. https://www.nhlbi.nih.gov/health/educational/lose_wt/risk.htm. Accessed February 21, 2022.

2. Hales C, Carroll M, Fryar C, Ogden C. Prevalence of Obesity and Severe Obesity Among Adults: United States, 2017–2018. www.cdc.gov. Published 2020.  https://www.cdc.gov/nchs/products/databriefs/db360...

3. Akil L, Ahmad HA. Relationships between Obesity and Cardiovascular Diseases in Four Southern States and Colorado.  Journal of Health Care for the Poor and Underserved. 2011;22(4A):61-72. doi:10.1353/hpu.2011.0166

4. Sterling S, Judd S, Bertrand B, Carson TL, Chandler-Laney P, Baskin ML. Dietary Patterns Among Overweight and Obese African-American Women Living in the Rural South.  Journal of Racial and Ethnic Health Disparities. 2017;5(1):141-150. doi:10.1007/s40615-017-0351-3

5. Ozodiegwu ID, Doctor HV, Quinn M, Mercer LD, Omoike OE, Mamudu HM. Is the positive association between middle-income and rich household wealth and adult sub-Saharan African women’s overweight status modified by the level of education attainment? A cross-sectional study of 22 countries.  BMC Public Health. 2020;20(1). doi:10.1186/s12889-020-08956-3