TBA – Psychology

Answer

Causal Obesity Model

Worldwide, obesity has consistently become a prominent health concern, emerging above the traditional health concerns. Dalvand, et al (2015) cited that obesity prevalence has risen immensely over the last century and has dramatically effectuated since the millennium, a phenomenon which has been highly linked with various shifts in lifestyle. Though there is a plethora of scientific research arguing against BMI as a good measure for overweight by itself, the measure remains the best predictor of abdominal fat while waist circumference has been shown to be the best predictor of visceral fat, so clinical research has evolved to a reinforced utilization of both in measuring obesity. Compounded with the existing issue of underweight, such concerns expose the double burden of disease and may put stress on rural healthcare (Little, et al., 2016). Since BMI and waist circumference are inherent correlates, some clinical researchers have used either to estimate obesity despite the debates. There is a myriad of lifestyle disorders related to obesity, with hypertension being the most common (Saliba, & Maffett, 2019). Age, sex, physical activity, and socioeconomic state have shown a relation to obesity (Dalvand, et al., 2015). The current paper seeks to examine the findings of a previous research on obesity and waist circumference among Iranian adults by Dalvand, et al. (2015). The causal relationship between age and waist circumference is examined. 

Method

Participants

Data for the present study were derived from the third round of the survey of Noncommunicable Diseases Risk Factors Surveillance in Iran. Participants were interviewed at their homes by trained healthcare workers from 43 medical schools and a blood sample was taken after receiving a verbal informed consent. After excluding pregnant women, the data analyzed included 18,990 women and men aged ≥20 years (Dalvand, et al., 2015).  This population-based cross-sectional study was conducted by Iran Center for Diseases Control. A cluster sampling design was used to produce representative data for that age range in Iran. The number of clusters in each province was proportional to the size of that province, each cluster comprising 10 men and 10 women. For each province, a total of 50 clusters including 20 participants, two males and two females in each 10-year age group, were selected using a proportional-to-size systematic sampling scheme.

Material

Waist circumference and obesity were treated as the main response variables of the study.

Age, Physical Activity, Smoking Status, Blood Pressure, Fasting Blood Glucose and Cholesterol were the independent variables.

Procedure

The households’ addresses were extracted by Iran’s Post Company. Eventually, participants were interviewed at their homes after receiving an informed consent by trained healthcare workers. Based on the STEP-wise approach of WHO, STEPS is a sequential process, starting with gathering information on key risk factors by the use of questionnaires, then moving to simple physical measurements, and only then recommending the collection of blood samples for biochemical assessment.

Data Analysis

Table 1: Model for first IV (Glucose).

Table 2: Model for second IV (Cholesterol).

Table 3: Table with the two IVs, comparing to null.

Table 4: Table with the two IVs, comparing to best model.

Obesity - Cholesterol

Fig 1: Cluster plot of Obesity vs Cholesterol

Obesity - Glucose

Fig 2: Cluster plot of Obesity vs Glucose.

There is indeed a correlation between Obesity and Glucose, a positive correlation – which indicates that as one’s glucose content rises, the obesity risk also rises. The same phenomenon is seen with cholesterol, with a higher correlation. As cholesterol content rises, obesity risk also rises. This is further seen in the causal models for both individual and covariate models.

For the first variable, Gluose, the probability of the model has increased from 50% to 58.1%, after observing the model, with a R² of 0.042.

For the second variable, Cholesterol, the probability of the model has increased from 50% to 100%, after observing the model, with a R² of 0.245.

The combination of both independent variables results in an improved model n comparison to null. The model’s probability increases from 33.3% to 84.8%, with an R² of 0.288. This covariate model highly improves. This is the same case when the covariate model is compared with the best model (the most probable or feasible model after observing the data) since we see the probabilities given the observed data being the highest.

Discussion

The findings of this study are consistent with previous research papers that have concluded that glucose and cholesterol contents are among the main factors influencing obesity. This study found significant correlation between each of the independent variables and the dependent variable (obesity). Therefore, it is conclusive, from the overserved data, that cholesterol and glucose contents of one’s body have a linear association with the person’s obesity.

Further, the study established a causal relationship between both cholesterol and glucose contents and obesity with positive coefficients. This affirms the previous findings that both high cholesterol and high glucose contents increase the risks of obesity. Essentially, a combination of both heighten the risk due to the higher combined effects.