Discriminating Bangladeshi Adults by Level of Obesity Volume 3 - Issue 1

Bhuyan KC¹* and Jannatul Fardus²

• ¹Professor (Retired) of Statistics, American International University, Bangladesh
• ² Lecturer in Statistics, American International University, Bangladesh

Received: January 29, 2019;   Published: February 05, 2019

*Corresponding author: KC Bhuyan, Professor (Retired) of Statistics, American International University, Bangladesh

DOI: 10.32474/LOJMS.2019.03.000152

Abstract PDF

Abstract

The present study was based on data collected from 900 respondents of both urban and rural areas of Bangladesh. Among them 44.3 percent were obese and severe obese. Majority (70.6 %) were diabetic patients With the increase in age significant increase in prevalence rate of obesity and severe obesity was observed. Higher prevalence rate was also observed among retired respondents. An upward shift of prevalence rate was observed among the higher educated respondents. Urban people were more obese and severe obese. Higher proportion of females was obese. This was true for housewives also. Obesity and severe obesity were prevalent among respondents who were not involved in any physical activity. The respondents of different levels of obesity were discriminated due to the variables residence, age, gender, and marital status. This was decided by the measurement of larger correlation coefficients of variables and discriminant scores.

Introduction

The prevalence of overweight and obesity has increased rapidly over the last decades especially in developed countries [1-3]. In 2011 WHO estimated that globally approximately 1.5 billion adults (age > 20 years) were overweight and more than 500 million adults were obese [4]. Obesity is generally associated to a significantly higher risk of arterial hypertension, diabetes mellitus (DM), hepatic steatosis, yperdyslipidemia and renal failure [5]. The major contribution of obesity is to lead the increase in the prevalence of chronic diseases and cancers [6-9]. The most common medical morbidities associated with obesity include impaired glucose tolerance and metabolic syndrome [10,11]. Behavioral factors have significant effects on metabolic risk. It had been observed in some research findings that youth who do not meet guidelines for dietary behavior, physical activity and sedentary behavior have greater insulin resistance than those who do meet guideline [12]. For this reasons, World Health Organization considers the epidemic a worldwide problem which requires public health intervention [13] that act on different factors associated with overweight and obesity as well as technological changes that have lowered the cost of living of the people so that people can avail sufficient food with required protein. Efforts are needed to improve the Economic, political, social and environmental conditions so that congenial atmosphere prevails in the society for maintaining healthy life of the people. But people are less aware, specially the rural people, of the disease and the factors responsible for the disease. Even government and public health planners remain largely unaware of the current prevalence of obesity and diabetes. As a result, the factors responsible for the disease are not well identified. The aim of this paper is to identify the socioeconomic factors responsible for obesity and severe obesity among diabetic and normal subjects of some rural and urban people in Bangladesh. The important factors for obesity and severe obesity will be identified by discriminant analysis, where largest correlation coefficient of any variable with discriminant function scores indicate the most important variable responsible for obesity.

Methodology

The study was based on data collected from both urban and rural people of Bangladesh. As per objective of the study the main target was to collect data from diabetic patients. The investigated diabetic patients were 544. To study the variability of socioeconomic variables for diabetic and non-diabetic people, some respondents were also investigated as a control group. The number of this later group of respondents was 346. However, among this later group of respondents there were 91 diabetic patients. Thus, finally, the analysis was performed using the data of 635 diabetic patients and 265 non-diabetic people. The data were collected through a predesigned and pre-tested questionnaire during the months of May and June, 2015 by some undergraduate and post graduate students of American International University-Bangladesh, most of whom were doctors and nurses of the department of Public Health of the university and they were associated with public health services. The data were collected from the diabetic patients of the working places of the investigators according to their convenience.

Data have also been collected from parents/guardians of 200 randomly selected students of different disciplines of the university, on the assumption that the respondents would be of normal group of people. But during investigation some of them were found as diabetic patients. However, from the filled-in questionnaires 356 were found in completed form and the information of these 356 respondents were included in the analysis. The questionnaire contained questions related to socio-demographic characters of each person. Questionnaire also contained questions related to the stage and type of diabetes, treatment stage of disease, pre-cautions against the disease and the stage of complications due to the disease. The latter information were provided by the diabetic patients. The information regarding blood sugar level and blood pressure level were also noted down according to the latest measurement by doctors/diagnostic centers.

Some of the variables observed were qualitative in character and some were quantitative. All variables were transformed to nominal form by assigning numbers to do the discriminant analysis. The variables included for the analysis were residence of the respondents, their age, gender, marital status, religion, level of education, occupation, type of work, monthly income, smoking habit and prevalence of diabetes. The analysis was done by using SPSS [ version 20.0]. The level of obesity was measured by BMI [weight in kg /height in (m) [2]. The respondents were classified as underweight [BMI < 20], overweight [BMI ,20 -25], obese [BMI < 30] and severe obese [ BMI 30+]. Discriminant analysis was done to identify the most important factors responsible for different levels of obesity. Besides the discriminant analysis, the association of different socioeconomic variables with level of obesity were investigated. Significant association was decided by chi-square test with p-value< 0.05.

Results

It was observed from the analysis that among 900 respondents 7.6 percent were underweight (Table 1) and 19.1 percent of them were from rural area. Maximum (43.1) of the respondents were overweight and 20.9 percent were from rural people. Severe obesity was observed among 15.3 percent people and obese respondents were 34.0 percent. In every case of level of obesity, the majority were from urban area. Of course, major respondents (81.4%) were from urban area. However, the differences in proportions of level of obesity according to residential area were not significant [P (χ2 ≥ 5.128) = 0.528]. The levels of obesity were similar for both urban and rural areas. The level of obesity were significantly different among males and females (Table 2), P (χ2 ≥ 27.546) = 0.000]. There were 58.9 percent males among the respondents and 47.2 percent were overweight. The corresponding percentage among females was 37.3. However, compared to males more females were severely obese. Obesity and severe obesity were observed almost similar among Muslims and Non-Muslims (Table 3). But more Muslim respondents (43.8%) were overweight compared to Non-Muslim respondents (38.8%). Significant differences in proportions of obesity among the two religious groups were noted [ P (χ2 ≥ 10.82) =0.012]. Among the investigated respondents 92.6 percent were currently married and 43.1 percent of them were overweight (Table 4). Similar overweight group was noted among the other group of respondents. However, there was significant differences in proportions of different levels of obesity among the two marital groups of respondents [P (χ2 ≥ 22.933) = 0.028] Majority (52.9%) of the respondents were of age 50 years and above and 48.5 percent of them were overweight (Table 5).

The overweight groups among the respondents of ages 25 – 40 and 40 – 50 were 35.5 and 36.6, respectively. Levels of obesity was significantly associated with levels of ages [ P (χ2 ≥ 18.34) =0.008] Most (62.6%) of the respondents were at least graduate and 40.2 percent alone were graduate. Among them 41.7 percent were overweight. The overweight group was (Table 6) higher among respondents of all levels of education. More overweight people was observed among illiterate respondents. The level of obesity and level of education were significantly associated [P (χ2 ≥ 26.376) = 0.034]. Higher proportion of respondents (23.2%, Table 7) were businessmen and 45.5 percent of them were overweight. Maximum overweight group (53.8%) was observed among agriculturists followed by retired persons (45.9%). The overall overweight group was maximum (43.1%). Maximum (25.5%) severe obesity was noted among housewives. The proportions of different levels of obesity according to professional variations were significant [P (χ2 ≥ 46.472) = 0.000]. The lower income (< 20,000.00 Taka) group of people were more (34.2%) and 48.4% of them were overweight (Table 8) More overweight group of people were observed (49.0%) among the respondents who had income 20,000.00 -< 30,000.00. This group of people were 20.2 percent.

The data indicated that 54.4% respondents had income less than 30,000.00 taka. More obese people was observed among them who had income 30,000.00 - < 40,000.00 taka followed by the group of people who had income 50, 000.00+. Severe obese group was also more (30.3% 0) among them. Significant association was noted between the level of obesity and the level of income [P (χ2 ≥ 64.994) = 0.00] It was noted that (Table 9) 50 percent respondents were involved in official work with or without physical labor. These two groups of people were 450 and 164 of them were obese. Again, those who were not doing any physical labor (23.1%) obesity and severe obesity was more prevalent among them (51.4%). However, level of obesity was not associated with type of work [P (χ2 ≥ 11.905) = 0.453]. So far we had discussed the results of association of levels of obesity and some socioeconomic characteristics. Now, let us observe the association of level of obesity and prevalence of diabetes. Table 10 showed that 67.6 percent underweight respondents were affected by diabetes. With the increase in levels of body mass index [BMI] the rates of prevalence of diabetes were also increased. However, the differentials in rates of obesity and rates of prevalence of diabetes were not significant as [P (χ2 ≥ 0.851) = 0.837]. (Table 10) In one study [18], it was reported that smoking is one of the factor to increase the level of obesity. The present study also indicated similar result (Table 11). Among the smokers 47.2 percent were overweight and 37.2 percent were obese. The corresponding figures among non-smokers were 41.3 and 32.5, respectively. The association between smoking habit and level of obesity was significant [ P (χ2 ≥ 20.189) = 0.0.002]. Among the respondents 31.3 percent were smokers (Table 11) and among the smokers 47.2 percent were overweight and 37.2 percent were obese.

The corresponding figures among non-smokers were 41.3 and 32.5, respectively. The association between smoking habit and level of obesity was significant [ P (χ2 ≥ 20.189) = 0.0.002]. Discriminant Analysis It was observed that levels of obesity were varied differently with the variation of different social factors. Thus, we were in search of identification of most important variables to discriminate the respondents according to various of levels of obesity. This was done by discriminant analysis. The analysis helps to identify the important variables for which the groups of respondents were significantly different [14]. The variables which were included in the analysis were sufficient to discriminate the different groups of respondents according to their level of obesity as Box’s M = 287.926 and the corresponding F= 1.403 with p – value = 0.000. The analysis provided 3 discriminant functions for 4 groups of respondents. The first function was significant as values of Wilk’s for first, second and third functions were 0.918, 0.973 and 0.994, respectively and the corresponding values were 75.920(p-value=0.0000, 24.493 (p-value=0.222) and 5.065 (p-value=0.829). The standardized canonical discriminant function coefficients were presented in (Table 12). From the discriminant analysis we had the correlation coefficients of variables and the discriminant functions scores. These coefficients were shown in (Table 13). The analysis indicated that the respondents of different levels of obesity were significantly different according to sociodemographic variables. The important variable for discrimination was residence followed by age. The other important variables were gender and marital status.

Discussion

The analysis presented here was done from the data collected from 635 diabetic patients and 265 control group of respondents. The respondents were investigated mostly by the doctors and nurses from their working places. The selection procedure was a convenient sampling plan. The investigated respondents were divided in to 4 groups according to their level of obesity, where levels of obesity were decided by their levels of BMI. Around 50 percent respondents were obese and severe obese. Higher (71.6%) prevalence rate of diabetes was noted among the obese and severe obese group of respondents. Similar finding was also noted in another study [16]. The prevalence of obesity and severe obesity were significantly associated with age, religion, education, occupation, marital status, income and smoking habit. Similar results were also noted in separate studies [ 5-9] [15-17]. Around 50.6 percent people of urban area were obese and severe obese. This result is also similar as was observed in another study [17]. Discriminant analysis also indicated that some of the socioeconomic variables were responsible for increased rates of obesity and severe obesity. These variables were residence, age, gender and marital status. Among these 4 variables residence and age were very important.

Conclusion

The analysis was done from the collected information of 900 respondents. They were classified as underweight (BMI <20), overweight (BMI ,20 - <25), obese (BMI, 25 - < 30) and severe obese (BMI = 30+). The percentages of these four group of respondents were 7.6, 43.1.34.0 and 15.3, respectively. Among the respondents overweight and obesity were more prevalent. The obesity is one of the risk factor of prevalence of non-communicable diseases [NCD] and it enhances the arterial hypertension, diabetes renal failure etc. [3]. In this study also, higher prevalence rate of diabetes was observed among them. Among the respondents 84.1 percent were of the age 40 years and above and 42.8 percent of them were obese and severe obese. Again, prevalence of diabetes was more among these groups. This finding is similar to that observed in both home and abroad [17-19]. The discriminant analysis showed that residence, age, gender and marital status were more important to differentiate among the respondents of different levels of obesity. The incidence of obesity cannot be avoided, but its prevalence can be reduced by implementing appropriate action plan. The following actions are very important to reduce the prevalence of obesity. These are:

a. Halt the rise in body weight by encouraging people so that they can take healthy homemade food.

b. People may be encouraged to do some sort of physical labor after or before official work. This is for the in service, private or government, people.

c. Counseling is needed for the obese children and adolescents.

d. To motivate people so that they become careful about the danger of obesity and its adverse effects of health. The public health authority can play a decisive role for the above steps.

References

1. Dhurandhar NV (2001) Contribution of pathogene in human obesity. Drug News & Perspectives 7(5): 307-313.
2. Berg C, Rosengren A, Aires N, Lappas G, Torén K, et al. (2005) Trends in overweight and obesity from 1985 to 2002 in Goteborg, west Sweden. Int J Obes (Lond) 29 (8): 916-924.
3. Skliros EA, Merkouris P, Sotiropoulos A, Xipnitos C, Liva H, et al. (2008) The relationship between body mass index and hypertension in elderly Greeks, the Nemea Primary Care Study. Journal of the American Geriatrics Society 56(5): 954-955.
4. WHO (2018) fact sheet n0 311 at Obesity and overweight.
5. Calamusa G, Amodio E, Costantino C, Pasquale MD (2012) Body mass index and factors associated with overweight and obesity : a crosssectional study of adult subjects living in a small city of Western Sicily ( Italy). Italian Journal of Public Health 9 (3): e7539-1-e7539- 8.
6. Prasad SB, Fahrtash F, Malaiapan Y, Meredith IT, Cameron J (2009) Obesity and the metabolic syndrome in patients with acute myocardial infraction. International Journal of Cardiology 144(3): 450-451.
7. Redon J1, Cifkova R, Laurent S, Nilsson P, Narkiewicz K, et al. (2008) The metabolic syndrome in in hypertension: European Society of hypertension position statement, Journal of Hypertension 26(10): 1891-1900.
8. Basen Engquist K, Chang M (2011) Obesity and cancer risk: recent review and evidence, Curr Oncol Rep 13(1): 71-76.
9. Bertino G, Ardiri AM, Alì FT, Boemi PM, Cilio D, et al. (2006) Obesity and related diseases : an epidemiologic study in eastern Sicily. Minerva Gastroenterol Dietol 52(4): 379-385.
10. Pulgaron ER (2013) Child obesity: a review of increased risk for physical and psychological combordities. ClinTher 35(1): 18-32.
11. Sinha R, Fisch G, Teague B, Tamborlane WV, Banyas B, et al. (2002) Prevalence of impaired glucose tolerance among children and adolescents with marked obesity. N Engl J Med 346(11): 802-810.
12. Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, et al. (2014) Obesity and the metabolic syndrome in children and adolescents. N Engl J Med 350(23): 2362-2374.
13. WHO (2007) The challenge of obesity in the WHO European region and the strategies for response, (Eds) Branca F, Nikogosion H, Lobstein T, Copenhagen WHO.
14. Farzana F, Bhuyan KC (2018) Identification of factors responsible for obesity in children and adolescents of some affluent families, Biomed J Sci & Tech 10(5): 1-9.
15. Bhuyan KC (2004) Multivariate Analysis and its Applications, New Central Book Agency(p) Ltd.
16. Bhuyan KC, Mortuza A, Fardus J (2018) Discriminating patients suffering from non-communicable diseases: A case study among Bangladeshi adults. Biomed J Sci & Tech 10(1): 1-7.
17. Fardus J, Bhuyan KC (20016) Discriminating diabetic patients of some urban and rural areas of Bangladesh: A discriminant analysis approach EMBJ 11(19): 134-140.
18. Akter S, Rahaman MM, Sarah Krull Abe, Sultana S (2014) Prevalence of diabetes and their risk factors among Bangladeshi adults: a nationwide survey. Bull World Health Organ 92(3): 204-213.
19. Sudeep S (2012) Silent killer, economic opportunity: Rethinking noncommunicable disease. Centre on Global Health Security, World Bank.