Physical Activity, Sedentary Behaviors, and BMI
Physical Activity, Sedentary Behaviors, and BMI
This study has addressed the concept of sedentary behavior and PA by looking specifically at how watching television and engaging in moderate levels of PA effect BMI classification. The effect of watching television more than doubled between the normal weight classification and the overweight classification. For those in the overweight classification, an extra 2 h spent watching television per day was associated with nearly one additional BMI point. But, this finding was not consistent across all BMI classifications. Watching television was only incrementally detrimental to BMI when viewed from the perspective of moving from normal weight to overweight. Among respondents who were already overweight, a similar effect was not observed. Conversely, the effect of moderate PA on BMI was greater for those in higher BMI classifications than for those in lower BMI classifications. Specifically, the transition from obese class I to obese class II demonstrated a significant decrease in BMI with increased moderate PA (F = 4.54, p = 0.033). This finding suggests there is much to be gained from engaging individuals with higher BMI in moderate PA.
Our findings about the heterogeneous effect of PA is somewhat supportive of Cooper and colleagues' finding that adults of differing BMI classifications also had differing levels of PA. Our finding supplements their study by isolating the statistical influence of differing levels of PA on BMI at different BMI classifications.
This study demonstrates the significant association of socio-demographics with BMI classification. Self-reporting as African-American had a significant effect on BMI classification which increased as BMI increased. For those in the overweight BMI classification, self-reporting as African-American was associated with nearly four additional BMI points. This effect was nearly four times as great as the effect of self-reporting as African-American in the normal weight BMI classification. Conversely, self-reporting as Hispanic was only significantly associated with lower BMI classifications such as normal weight, overweight, and obese class I. However, the effect of self-reporting as Hispanic was homogenous across these categories.
Age was significantly associated with all BMI classifications. Among higher levels of BMI classification, the effect of increasing age became stronger. For those in the obese class III BMI classification, each additional year aged was associated with a reduction of one quarter of a BMI point. Self-reporting as male was also significantly associated with lower BMI among the two lowest BMI classifications, normal weight and overweight. However, among the highest BMI classification, obese class III, self-reporting as female was significantly associated with higher BMI. The difference in the effect of being female on BMI was also significantly higher for those in the obese class III BMI classification versus the obese class II BMI classification. Self-reporting as female among the obese class III BMI classification was associated with nearly 3 additional BMI points. Educational attainment was also significantly related to BMI among the higher BMI classifications; obese class I, obese class II, and obese class III. For every additional year of education attained among those in the obese class I BMI classification, the effect on BMI was a reduction of a third of a BMI point. Finally, there did not appear to be significant results to report based on participants' residential rurality. It appeared this variable was of less importance than other socio-demographic variables.
A noteworthy finding in this study is the effect moderate PA had on BMI among those in the "obese class III" BMI classification, also known as the "morbidly obese". Part of the challenge in addressing PA for the morbidly obese is overcoming barriers endemic to physical conditions of obese individuals. One such barrier is the perception of breathlessness associated with exercise. This barrier can prevent obese individuals from beginning an exercise regimen. However, there is encouraging evidence that training programs incorporating respiratory muscle development can be successful in helping obese individuals reach their PA goals, improve their metabolic health, and sustain their level of exercise. In addition, mobility limitations of obese individuals caused by diseases such as osteoarthritis and joint pain, must be taken into account when planning appropriate PA programs. Another barrier to PA is motivation, particularly because obese individuals often report a lack of energy or feeling too tired to exercise. These are all unique barriers to PA that programs must recognize in order to be successful.
In addition to indicating the heterogeneous effect of sedentary behaviors among BMI classifications, our study also highlights the role of socio-demographic factors among the various degrees of obesity. While the coefficients in our study identified the unique effect of each independent variable on BMI, the results indicate future research could benefit by examining interaction effects of sedentary and lifestyle behaviors with select demographic groups such as Hispanics and African-Americans. Such analysis could be a powerful complement to the broader findings of our study, and likely highlight the need to tailor interventions not only toward individual BMI classifications, but also toward the multi-cultural composition of those classifications. Addressing the unique needs of the Hispanic and African-American populations can improve the chances of developing programs that successfully improve the health of these populations.
Our study used a unique statistical method to analyze the homogeneity of effect between sedentary behaviors, socio-demographics, and BMI classification. However, simultaneous quantile regression does not have a true equivalent of the coefficient of determination which makes it difficult to establish how much of the variance in the dependent variable was explained with the model. Further, while chronic diseases were recognized in the selection criteria, individual chronic disease characteristics were not analyzed in this study. Future studies would benefit from incorporating individual disease variables as well as individual values of lipoprotein lipase (LPL), low-density lipoprotein cholesterol (LDL-C), blood sugar, and other biometric markers where appropriate. Finally, the data analyzed in this study was self-reported from several communities in central Texas. As a result, it has limited generalizability. It must also be noted the BVHA solicited households for individual responses to the study instrument. As such, individuals not living in households are likely not well represented in the survey results.
Discussion
The Effects of PA and Sedentary Behavior on BMI Classification are not Homogeneous
This study has addressed the concept of sedentary behavior and PA by looking specifically at how watching television and engaging in moderate levels of PA effect BMI classification. The effect of watching television more than doubled between the normal weight classification and the overweight classification. For those in the overweight classification, an extra 2 h spent watching television per day was associated with nearly one additional BMI point. But, this finding was not consistent across all BMI classifications. Watching television was only incrementally detrimental to BMI when viewed from the perspective of moving from normal weight to overweight. Among respondents who were already overweight, a similar effect was not observed. Conversely, the effect of moderate PA on BMI was greater for those in higher BMI classifications than for those in lower BMI classifications. Specifically, the transition from obese class I to obese class II demonstrated a significant decrease in BMI with increased moderate PA (F = 4.54, p = 0.033). This finding suggests there is much to be gained from engaging individuals with higher BMI in moderate PA.
Our findings about the heterogeneous effect of PA is somewhat supportive of Cooper and colleagues' finding that adults of differing BMI classifications also had differing levels of PA. Our finding supplements their study by isolating the statistical influence of differing levels of PA on BMI at different BMI classifications.
Socio-demographic Effects on BMI Classification are Noteworthy
This study demonstrates the significant association of socio-demographics with BMI classification. Self-reporting as African-American had a significant effect on BMI classification which increased as BMI increased. For those in the overweight BMI classification, self-reporting as African-American was associated with nearly four additional BMI points. This effect was nearly four times as great as the effect of self-reporting as African-American in the normal weight BMI classification. Conversely, self-reporting as Hispanic was only significantly associated with lower BMI classifications such as normal weight, overweight, and obese class I. However, the effect of self-reporting as Hispanic was homogenous across these categories.
Age was significantly associated with all BMI classifications. Among higher levels of BMI classification, the effect of increasing age became stronger. For those in the obese class III BMI classification, each additional year aged was associated with a reduction of one quarter of a BMI point. Self-reporting as male was also significantly associated with lower BMI among the two lowest BMI classifications, normal weight and overweight. However, among the highest BMI classification, obese class III, self-reporting as female was significantly associated with higher BMI. The difference in the effect of being female on BMI was also significantly higher for those in the obese class III BMI classification versus the obese class II BMI classification. Self-reporting as female among the obese class III BMI classification was associated with nearly 3 additional BMI points. Educational attainment was also significantly related to BMI among the higher BMI classifications; obese class I, obese class II, and obese class III. For every additional year of education attained among those in the obese class I BMI classification, the effect on BMI was a reduction of a third of a BMI point. Finally, there did not appear to be significant results to report based on participants' residential rurality. It appeared this variable was of less importance than other socio-demographic variables.
Is it Time to Exercise?
A noteworthy finding in this study is the effect moderate PA had on BMI among those in the "obese class III" BMI classification, also known as the "morbidly obese". Part of the challenge in addressing PA for the morbidly obese is overcoming barriers endemic to physical conditions of obese individuals. One such barrier is the perception of breathlessness associated with exercise. This barrier can prevent obese individuals from beginning an exercise regimen. However, there is encouraging evidence that training programs incorporating respiratory muscle development can be successful in helping obese individuals reach their PA goals, improve their metabolic health, and sustain their level of exercise. In addition, mobility limitations of obese individuals caused by diseases such as osteoarthritis and joint pain, must be taken into account when planning appropriate PA programs. Another barrier to PA is motivation, particularly because obese individuals often report a lack of energy or feeling too tired to exercise. These are all unique barriers to PA that programs must recognize in order to be successful.
Targeted Interventions are Needed
In addition to indicating the heterogeneous effect of sedentary behaviors among BMI classifications, our study also highlights the role of socio-demographic factors among the various degrees of obesity. While the coefficients in our study identified the unique effect of each independent variable on BMI, the results indicate future research could benefit by examining interaction effects of sedentary and lifestyle behaviors with select demographic groups such as Hispanics and African-Americans. Such analysis could be a powerful complement to the broader findings of our study, and likely highlight the need to tailor interventions not only toward individual BMI classifications, but also toward the multi-cultural composition of those classifications. Addressing the unique needs of the Hispanic and African-American populations can improve the chances of developing programs that successfully improve the health of these populations.
Limitations
Our study used a unique statistical method to analyze the homogeneity of effect between sedentary behaviors, socio-demographics, and BMI classification. However, simultaneous quantile regression does not have a true equivalent of the coefficient of determination which makes it difficult to establish how much of the variance in the dependent variable was explained with the model. Further, while chronic diseases were recognized in the selection criteria, individual chronic disease characteristics were not analyzed in this study. Future studies would benefit from incorporating individual disease variables as well as individual values of lipoprotein lipase (LPL), low-density lipoprotein cholesterol (LDL-C), blood sugar, and other biometric markers where appropriate. Finally, the data analyzed in this study was self-reported from several communities in central Texas. As a result, it has limited generalizability. It must also be noted the BVHA solicited households for individual responses to the study instrument. As such, individuals not living in households are likely not well represented in the survey results.