A study by (Nyrnes, Jorde & Vining, 2006) on body mass index (BMI) and serum thyroid- stimulating Hormone is ideally suited as a study on a health-related concern. This particular study was conducted to find out whether BMI and TSH correlate with one another. Measurements of heights and weights were taken to calculate the BMI as the response variable (dependent variable). For independent variables, serum TSH and smoking status were recorded.

Moreover, age and gender were included as variables of interest (covariables). The response variable, in this case, is a continuous variable while the regressor variables (independent variables) have different levels of measurements. Gender and smoking status are dichotomous variables while TSH and age are continuous variables.

Multiple linear regression analysis was conducted due to the nature of the data model; because we have more than one regressor (Montgomery, Peck & Vining, 2012). LinearLinear regression was conducted between gender and smoking status on the BMI, to investigate if there is an association between gender and smoking status of the participants regarding BMI. Moreover, gender-specific linear regressions were performed for both smokers and non-smokers to study sex-specific correlation with BMI. The effect of age was analyzed where only smokers of age 49 years and above were considered to explore the association of BMI and TSH. The results of the analysis show a positive and significant correlation between TSH and BMI for non-smoking participants; both female and male (Nyrnes et al., 2006).

The correlation between BMI and TSH does not imply that the analysis is causal (i.e., a change in TSH does not necessarily cause a change in the BMI). As much as a significant positive relationship is shown between the two (i.e., as TSH increased, BMI also increased), there could be other confounding variables that affect both BMI and TSH. Causation implies effect or cause, where an occurrence results in another occurrence (Altman & Krzywinski, 2015). Causation is more evident in controlled experiments and less precise for observational studies. Therefore, from this study, we can only report for correlation and not causation.

References

Altman, N., & Krzywinski, M. (2015). Points of Significance: Association, correlation, and causation.

Montgomery, D. C., Peck, E. A., & Vining, G. G. (2012). Introduction to linear regression analysis (Vol. 821). John Wiley & Sons.

Nyrnes, A., Jorde, R., & Sundsfjord, J. (2006). Serum TSH is positively associated with BMI. International journal of obesity, 30(1), 100.