CMHL1001 Evidence Informed Health Practice For Body Mass Index

Answer:

Shobha, R., Prajakta, P.W. (2013) Performance of Waist Circumference Relative to BMI in Predicting Risk of Obesity and Hypertension among Affluent Indian Adults. Biometry and Nutrition Unit, Agharkar Research Institute, Pune, India. Vol.5, No.8A3, 16-22 (2013) Health https://dx.doi.org/10.4236/health.2013.58A3003 

Body Mass Index as a measure of adiposity is more reliable in determining and predicting the risk of obesity and hypertension as compared to Waist Circumference (WC) and Waist Hip to Ratio (WHR). 

Waist Circumference (WC) and Waist Hip to Ratio as measures of adiposity are more reliable determiners in predicting the risk of obesity and hypertension particularly among Indians and other Asian natives than Body Mass Index (BMI). 

The researchers presented sufficient statistical evidence presented before that BMI does not reflect the mass of abdominal fat in all populations. Their study is justified also because there is evidence that at a certain common BMI, Indians among other Asian natives have greater amounts of abdominal fat including a larger visceral fat area than Europeans. This population is thus at higher risk of developing metabolic disorders among other cardiovascular diseases like Diabetes type 2 and/or hypertension as compared to other white Caucasian populations. Their study is justified as they sought find a more accurate and tailor-made approach to determine the risk of obesity and hypertension in this population. Among the variables that needed to be tested thus included Body Mass Index (BMI), Waist Circumference (WC) and Waist Hip Ratio (WHR). Another justification is that any predictive power of a given anthropometric index depends on the population under study and usua

Advantages of the Study

The use of the cross sectional study design choice was appropriate, cheap and less time-consuming making it possible to obtain and analyse data very fast.

The study design used captured several and/or multiple variables at the same time (DBP, SBP, WC, WHR, and BMI) and this makes comparison easier in research methods. Even further, research outcomes obtained were suitable in achieving the study purpose. 

Disadvantages of the Study

The study did not consider potential change in the variables among participants with time as it should have included the observation of changes in these variables over a period of time to make more informed conclusions.

The study used an opportunistic sampling procedure for participants particularly members of the Rotary Club who attended health camps and this might have biased the data obtained in the study. 

A need for gender equality in choosing participants

Seeking for informed consent from the participants

Confidentiality on personal information of all the participants including socio-economic information, blood pressure, among other measurements. 

The authors took appropriate measures in addressing crucial ethics issues in this study. In regard to the need for consent before taking part in any given study, it is indicated that all the participants gave their consent voluntarily. This was an informed consent because the researchers provided full information on their intended study during the health camps at the Agharkar Research Institute. Gender equality is also an important ethics issue to consider in research studies(Szklo & Nieto 2007). The researchers ensured gender balance in choosing participants despite the fact that the number of men was only more than that of women by 1. Another concern was a need for regulated research studies in regard to protocol. It is indicated that the researchers presented their study protocol to the institutional Re- search Advisory Committee at Agharkar Research Institute which approved it and therefore rendering legit and ethical and appropriate. However, the researchers did not indicate if they kept confidential, personal information on socio-economic and anthropometric for the participants in the study. 

The participants in the study included 419 members of different Rotary Clubs within Pune city-Maharashtra, India. These included 210 men and 209 women with their ages ranging between 30 and 60 years old. The researchers organized health camps within the Agharkar Research Institute located in Pune City on weekends during which they informed the potential participants on the proposed assessment. Each participant who agreed to voluntarily take part in the study gave their written consent to the researchers after which they were considered legible. 

The researchers used opportunity sampling as the main sampling method in their study. They included participants who could avail themselves at the health camps within the Agharkar Research Institute and particularly those who were willing to take part in the study. The approach is only advantageous as it is quick and easy in regard to choosing participants (Szklo & Nieto 2007). However, its disadvantage is that it might at times not give a sample that is representative of the whole population. Further, opportunity sampling can even be biased. 

This is a Level III-3 Evidence classification study 

Yes, I think the study design; in particular the cross sectional study was particularly suitable to achieve the purpose of the study. A cross sectional study was suitable first because first, it was used to either prove and/or disapprove general assumptions that WC can better predict the risk of cardiovascular diseases as compared to BMI in Asians. Cross sectional studies are also usually cheap and less time-consuming but still capture several and/or multiple variables at the same time. For instance, it is clear that this single study assessed multiple variables at a go including DBP, SBP, WC, WHR, BMI, and AUC among others. The outcomes obtained were very suitable in achieving the study purpose effectively and efficiently due to the use of the cross sectional design. 

A longitudinal study design could have been used in this particular research. In this regard, the researchers could obtain information from the selected sample on WC, WHR, and BMI for use in the final analysis. The data on the measurable variable above should be taken continuously a longer period of time and not just once as done in this research. The period of longitudinal observation should thus be for instance one year, in order to also detect cause and effect in regard to variations among men and women in DBP, SBP, WC, WHR and BMI. This longitudinal study will thus go beyond just the single moment as done in the cross sectional design to establish crucial events and/or sequences in regard to adiposity and the risk of cardiovascular diseases in India and Asians by extension. 

Among men, it is evident that the correlation of waist circumference (WC) with systolic blood pressure (SBP) increased particularly from the lower to higher tertile and this correlation was significant. Even so, it was a lower significance as compared to the correlation coefficient for WC with SBP (0.232) among women. The relationship between the WC with SBP among men can thus be said to be not as strong as that among women.

Among women, the correlation of WHR with DBP increased particularly from the lower to higher tertile and this correlation was significant (P<0.01). Even so, it was a higher significance as compared to the correlation coefficient for WHR with DBP among men. The relationship between the WHR with DBP among women can thus be said to be stronger as that among men. 

Type 1 statistical error: The waist circumference is not a sensitive and reliable predictor of the risk of hypertension. In essence, type 1 error is rejecting the null hypothesis falsely while it is true.

Type 2 Statistical error: This could be failing to reject the null hypothesis that waist circumference is a sensitive and reliable predictor of the risks of hypertension, when it is true that the waist circumference is not a sensitive and reliable predictor of the risk of hypertension.

The lower tertile of MBI, WC and WHR measures was used as the reference category in the computation of the hypertension odds ratios. The authors have indicated that this is the case in the interpretation part of the table under results. 

The operational definition of hypertension was that it referred to the blood pressure equal and/or above 140/90 mmHg. By extension, individuals were considered hypertensive if they were on anti-hypertensive medication. According to the authors, the Systolic and diastolic blood pressure were obtained when study participants were sitting. The OMRON Blood Pressure monitor used was put on the left arm after each participant had rested for 10 minutes. The researchers used an adult- sized cuff to help obtain accurate blood pressure measures. For each particular blood pressure variable (DBP or SBP), the researcher’s recorded two readings but the second reading was the one they used for the final analysis. 

Normal weighing individuals were those whose BMI was not overweight (23.0 Kg/M²) and/or obese (25.0 Kg/M²). From the Table 1, the number of men in count who were normal weight is calculated as below;

Number of Men under study N= 210

Percentage of overweight men 23.0 Kg/M² =22.9%

Number of overweight men of BMI of 23.0 Kg/M² 

=48.09

=49 overweight men

Number of obese men=57.1

=119.91

= 120 obese men.

Number of normal weight men=N-(No. of obese men + No. of overweight men)

=210-169

=41 Normal weighing men. 

Overweight individuals are those whose BMI ranged falls above 25.0 Kg/M² 

From the Table 1, the number of Females in count who was normal weight is calculated as below;

Number of Females under study N= 209

Percentage of overweight females=65.1%

Number of overweight females i.e. BMI of above 25.0 Kg/M²

=136.059

=137

Therefore 137 Females out of the total of 209 under study measured over weight. 

Upper tertile (obese men) total=120.

Lower tertile (normal weighing) no of men 41

Total men lower and upper=161

OR= (ad)/ (bc)

= (49×29)/(71×12)

OR=1.67

This odds ration indicates that there is a strong relationship between BMI and the prevalence of hypertension in the Asian population. 

The statement “the correlation coefficient of BMI with WC was much higher than with WHR in both men and women” as stated by the authors imply that it is more reliable to define abdominal obesity using WC other than using WHR. This is because when the two variables were compared statistically, it was evident that WC was closely associated with general adiposity than WHR, particularly when it was defined based on a BMI measure among adult Indians in Pune, India. Statistically, the authors established that CW correlates more closely with BMI (coefficient of 0.852 in men and 0.825 in women) as compared to the correlation of WHR with BMI (coefficient 0.331 in men and 0.251 in women. In this regard, it is more reliable to measure adiposity and its risks to cardiovascular diseases using WC as the determinant as compared to using WHR. 

 In table 3 ‘p<0.01’ indicate statistical significance of a correlation between two variables. For instance it has been used to denote the statistical significance of the correlation of WC with SBP in women. 

The high odds ratio (8.24) shows that there is a high risk of developing and/or prevalence of hypertension for individuals with high BMI, WHR and WC. On the other hand, the low odds ratio (0.81) shows that there is there a lower risk of developing and/or prevalence of hypertension for individuals with high BMI, WHR and WC. The 95% confidence interval (CI) for OR (8.24(2.67-25.43) which is larger than OR(0.81(0.39-1.68) shows that the later OR is more precise that the earlier. In regard to significance, the OR (8.24(2.67-25.43) which shows a stronger relationship between exposure and outcome, is statistically significant (p<0.01) as compared to that of the OR (0.81(0.39-1.68) which is not statistically significant (p<0.07 

Findings

The main and first finding that the researchers made was that WC, BMI and WHR and BP are among the predictors of hypertension. Even so, the extent to which one of these factors predicted the risk of hypertension varied with each variable. Secondly, they established that in comparison, women have a higher abdominal obesity (35.9%) than men (32.4%). In regard to the, age adjusted mean systolic Blood Pressure, women had (124.3 ± 15.1 mmHg) while men had (114.7 ± 17.2 mmHg). On the other hand, the mean diastolic Blood Pressure for women was 82.9 ± 10.5 mmHg while that of men was 76.4 ± 10.7 mmHg. Another finding was that hypertension prevalence was higher in men (34.3%) than in women (21.5%). The researchers also found out that despite the fact that WC and WHR measures increased in proportion with BMI increase, the Pearson’s correlation co- efficient of particularly BMI with Waist Circumference was more significant (p < 0.01) and higher in comparison with WHR among women and men. As a result, the researchers found out that WC can be a better technique in the assessment of adiposity and the risk of cardiovascular diseases. The researchers also used the odds ratio to show the variation in the relationship between WC or WHR and hypertension. They established that in men within the highest tertile of WC, the odds ratio (OR (95% confident interval) = 2.19 (1.08 - 4.45)] particularly for hypertension turned out significant i.e. p < 0.05. Odds ratio in men was also significant (p < 0.01) in the middle WC tertile. For women, the odds ratio of [8.24 (2.67 - 25.43) was obtained in the higher tertile while 5.71 (1.82 - 17.89) was obtained for the middle tertile. In regard to the optimal cutoff point for obesity and/or hypertension identification showed a significant area under the curve curve (p<0.01) and also sensitivity for Waist Circumference than the Waist to Hip Ratio (WHR) for both sexes. The researchers established further that the cutoffs that determine the risk of hypertension among other cardiovascular conditions were way lower as those suggested by the WHO especially among men. The cutoffs however compared to other related reports in Asian studies. The researchers concluded that WC could perform better as compared to both WHR and BMI in assessing hypertension and that it can thus be utilized in epidemiological studies as a convenient inexpensive and simple screening tool. I think the results are valid and I trust them because it used standard measurements and the research was overseen by the institutional committee. 

References

Shobha, R., Prajakta, P.W. (2013) Performance of Waist Circumference Relative to BMI in Predicting Risk of Obesity and Hypertension among Affluent Indian Adults. Biometry and Nutrition Unit, Agharkar Research Institute, Pune, India. Vol.5, No.8A3, 16-22 (2013) Health https://dx.doi.org/10.4236/health.2013.58A3003

Centre for Statistics in Medicine BMJ Statistics Notes. (Undated) Retrieved June 15, 2010. URL:https://www.csm-oxford.org.uk/index.aspx?o=1292.

Greeenfield B, Henry M, Weiss M, Tse SM, Guile JM, Dougherty G, Zhang X, Fombonne E, Lis E, Lapalme-Remis, Harnden B. (2008)Previously suicidal adolescents: Predictors of six-month outcome. Journal of the Canadian Association of Child and Adolescent Psychiatry. 17(4):197–201.[PMC free article] [PubMed]

Szklo M, Nieto FJ. (2007) Epidemiology: Beyond the basics. 2nd edition. Sudbury, MA: Jones and Bartlett Publishers.