Abstract
The prevalence of central obesity has been increasing rapidly in recent decades. Central obesity, measured by waist circumference, is the most dangerous type of obesity since it is closely related to chronic diseases, metabolic complications, and high COVID-19 infection rates. The objective of this study was to identify the dominant factor of central obesity among the adult population. The study used secondary data from a 2017 cross-sectional study conducted at Bojong Gede Public Health Center, Bogor Regency, Indonesia. A total of 85 men and women aged 25–64 years old were selected through purposive sampling and included in the analysis. The association between risk factors and central obesity were measured through chi-square bivariate analysis and multiple logistic regression multivariate analysis using IBM SPSS application version 22. The prevalence of central obesity was 70.6%. The results showed that sex (women), total blood cholesterol level (hypercholesterolemia), energy, protein, fat, and carbohydrate intake (>110%personal nutritional needs) were significantly associated with central obesity (p-value < 0.05). Hypercholesterolemia was the most dominant risk factor for central obesity (p-value = 0.032; OR = 4.21; 95%CI = 1.131–15.667) adjusted for confounders.
Introduction
Central obesity is a pandemic that puts a significant burden on global health [1]. Associated with an increase in fasting plasma triglycerides, LDL cholesterol, blood sugar and insulin, as well as a decrease in HDL cholesterol, central obesity is thought to be a major contributor in cardiovascular diseases, the leading cause of death in the past decade [2]. Central obesity is also associated with non-alcoholic fatty liver disease and several types of cancer (e.g. prostate and colorectal cancer) [3]. Current evidence suggests that people with central obesity and its complications are more vulnerable to infection by COVID-19 with worse symptoms and prognosis, leading to higher death rates [4].
WHO claims central obesity as one of the most difficult health problems to address, as its prevalence has been increasing rapidly in recent decades [5]. More than half of the adult population from various countries and regions experience central obesity, e.g. the United States (56.0%) [6], Southern Asia (68.9%) [7], and Mexico (74,0%) [8]. In two decades, the prevalence of central obesity in China has doubled in size, from 19.94%in 1989 to 43.15%in 2011 [9]. The increasing trend of central obesity has been observed in Indonesia as well. Data from Indonesia Basic Health Research showed that central obesity prevalence of the population aged ≥15 years old has increased from 18.8%in 2007 to 31.0%in 2018 [10]. Moreover, the prevalence of central obesity among adults aged ≥25 years old in Bogor, West Java, was found to be 51.3%[11]. A cohort study conducted among adults in Bogor from 2011 to 2018 showed that the incidence of central obesity was 55.4%with a hazard rate of 10 cases per 1000 person-years [12].
Central obesity is defined as a condition of excessive abdominal adiposity (both visceral and subcutaneous fat), which can be measured by waist circumference (WC) [13]. This condition happens as a result of positive energy balance and adipose tissue redistribution. Socio-demographical characteristics, health-related conditions, and behavior contribute to central obesity. Socio-demographical factors consist of aging [3], female sex [9], being married [14], low educational level [15], poor nutrition knowledge [16], and unemployment [17]. Health-related factors consist of high blood cholesterol level (known as hypercholesterolemia) [18] and moderate-to-high psychological stress level [17]. Behavioral factors consist of excessive energy and macronutrients intake [19], low dietary fiber intake [20], excessive cholesterol intake [21], low physical activity level [15], and smoking habit [22].
The rise in central obesity prevalence is a potential threat to the adult productivity level since it causes a large health and economic burden. In fact, central obesity is preventable as in most cases the risk factors are modifiable [23]. Previous research found the prevalence of central obesity in adult patients at Bojong Gede Public Health Center, Bogor Regency reached 70.6%[24]. Data on risk factors are available, but have not been further analyzed. Therefore, this study aimed to analyze the dominant risk factor associated with central obesity among adult patients at Bojong Gede Public Health Center, Bogor Regency, West Java Province, Indonesia in 2017.
Methods
Data source
Secondary data from Widiartha’s cross-sectional quantitative research was used as a data source of the current study [24]. The data were collected through purposive sampling in June –September 2017 at Bojong Gede Public Health Center, Bogor Regency, West Java Province, Indonesia. Male and female patients aged 18 to 64 years who underwent WC measurement were included as the study subjects. Pregnant or breastfeeding women, women using hormonal contraceptives, as well as subjects with complications, acute infections, heart attacks, strokes, or undergoing surgery less than 2 months before data collection were excluded from the study. This study comprised of 85 participants, all of whom provided complete information for any of the analyzed variables.
Data collection
Data collection was conducted by 9 trained enumerators with a background in nutrition and nursing education. Central obesity was determined using measures of WC based on WHO cut-offs for the Asian Region (men >90 cm; women >80 cm) [25]. The WC data were obtained from an average of 2 times measurement using measuring tape of 0.1 cm accuracy. As supporting data, body mass index (BMI) was measured and categorized according to WHO cut-offs for the Asia Pacific Region (BMI ≥25 kg/m2 defined as general obesity) [26]. Bodyweight and height was measured using a digital scale and microtoise. All anthropometric measures were assessed using international protocols.
Participants’ identity (birth date, sex, marital status, employment status, education level, and smoking habit) and nutrition knowledge were assessed through interviews using validated questionnaires. Non-fasting total blood cholesterol examination was conducted with standard procedures using the EasyTouch GCU Cholesterol Meter device. Stress level data were collected through interviews using a Perceived Stress Scale 10 (PSS 10) questionnaire. Energy and nutrient intake data were assessed using a semi-quantitative food frequency questionnaire (SQFFQ) with a span of one year, converted to kilocalories, grams, and milligrams with a nutrient content-analyzer software. Energy and nutrient intake data were compared to the calculation of individual energy and nutritional requirements, which calculated on basal metabolic rate, physical activity, age, and bodyweight factors [27]. Individual nutrition requirements for protein, fat, and carbohydrate were each determined at the level of 15%, 25%, and 60%of estimated energy requirement. Data on physical activity levels was obtained using the Global Physical Activity Questionnaire (GPAQ) in MET minutes/week.
Statistical analysis
The classification of each research variable is shown in Table 1. Categorical data were expressed in percentages. Bivariate analysis done with chi-square test, while multivariate analysis done with multiple logistic regression test. Test results with p-value <0.05 for bivariate and multivariate analysis were considered statistically significant. Statistical analysis was performed using the IBM SPSS application version 22.
Proportions of Central Obesity Based On Risk Factor Variables among Adult Patients at Bojong Gede Public Health Center, Bogor Regency, Indonesia in 2017
Proportions of Central Obesity Based On Risk Factor Variables among Adult Patients at Bojong Gede Public Health Center, Bogor Regency, Indonesia in 2017
Note: physical activity level unit is MET minutes/week; * = significant relationship (p-value < 0.05).
Characteristics of participants
The prevalence of central obesity was higher than general obesity among 85 adult patients at Bojong Gede Public Health Center (70.6%vs. 69.4%). The average WC was 88.89±13.59 cm. The average WC of women participants was substantially higher than men (89.89±13.12 cm vs. 83.84±15.36 cm). Table 1 shows the proportions of central obesity according to risk factor variables.
Most of the participants were >40 years old (69.4%), female (83.5%), married (90.6%), had high education level (61.2%), had poor nutrition knowledge (54.1%), dan unemployed (61.2%). The average score of nutrition knowledge was 6.58±2.08 (scale 1–9). The most commonly known nutritional information was the benefits of food consumption for body functions (94.1%answered correctly), while the least known were the recommended physical activity frequency (51.8%answered correctly) and nutrient compositions in a full meal (61.2%answered correctly). Among sociodemographic risk factors, sex is the only variable that had significant relationship (p-value <0.05) with central obesity, where central obesity was more common in women (OR = 6.188; 95%CI = 1.814–21.101).
It is known that more than half of the participants experience hypercholesterolemia (62,4%) and low psychological stress levels (57,6%). The median of the total blood cholesterol level was 213,00±79,92 mg/dL. There was a significant relationship between total blood cholesterol levels and central obesity. Central obesity was more commonly found in patients with hypercholesterolemia (OR = 2,970; 95%CI = 1,133–7,784).
Associations between behavioral risk factors and central obesity
Among the participants, energy, protein, fat and carbohydrate intake exceed 110%individual nutritional requirements (83,5%; 65,9%; 80,6%; dan 62,4%), low dietary fiber intake (78,8%), sufficient cholesterol intake (54,1%), physically active (83,5%), and do not smoke (91,8%) were the common behavioral characteristics. In average, protein intake contributes 12,4±2,14%; fat intake contributes 37,7±9,47%; and carbohydrate intake contributes 48,4±9,76%to total energy intake (TEI). A significant relationship was found between energy and macronutrient intake with central obesity. The odds of central obesity were higher for participants with excessive energy (OR = 6,188; 95%CI = 1,814–21,101), protein (OR = 4,929; 95%CI = 1,815–13,385), fat (OR = 9,158; 95%CI = 1,133–7,784) and carbohydrate intake (OR = 2,970; 95%CI = 1,133–7,784). No significant relationship was found between other behavioral factors and central obesity.
Associations after multivariate regression modelling
Selected variables with p-value <0,25 for the chi-square test (except for energy intake) were included in multivariate analysis. The energy intake variable might cause structural multicollinearity in the regression model since it was formed from the conversion of protein, fat, and carbohydrate intake variable. Multicollinearity is undesirable because it increases the standard error value, hence confound the results of data analysis. Table 2 showed hypercholesterolemia as the dominant factor related to central obesity (OR = 4.210; 95%CI = 1,131–15,667) after adjusted for sex, education level, nutrition knowledge, employment status, protein, fat, carbohydrate, dietary fiber, and cholesterol intake, and smoking habit variables.
Early and Final Model of Multiple Logistic Regression Analysis
Early and Final Model of Multiple Logistic Regression Analysis
*= significant relationship (p-value < 0.05).
The difference in the prevalence of the two types of obesity found among adult patients at Bojong Gede Public Health Center showed that central obesity can develop independently from general obesity (70.6%vs. 69.4%). This phenomenon, called normal-weight central obesity, is associated with a higher risk of mortality from cardiovascular diseases compared to obese individuals based on BMI indicators [25]. The prevalence found in this study is higher than the national (Indonesia) [15] and local (West Java) [11] central obesity prevalence, but not much different from the prevalence found in South Asian [7] and Mexican adults [8] as reported in national scale surveys.
Generally, the risk of developing central obesity increases with age, notably at 35 years and older as a consequence of increased fat deposits in the abdominal area [3, 14]. In this study, the mean WC of ‘>40 years old’ age group was slightly higher than ‘≤40 years old’ (89.05±12.72 cm vs. 88.53±15.68 cm), but the difference was insignificant. This finding reinforced the notion that there is an increasing trend of over nutrition at early adulthood [9].
The odds of central obesity were significantly higher for women than men (OR = 6,188) [9, 12]. The differences in the amount of fat reserves, hormonal functions, genetic features, and physical activity patterns between both sexes explain the higher risk of central obesity for women [28]. Considering that the majority of women participants were married, and the average age (47.1 years old) was in the range of menopausal age of Asian populations (42.1–49.5 years old) [29], higher prevalence of central obesity in women may also be affected by the experienced pregnancy, childbirth, and menopause [30].
In contrast to the previous study, no significant relationship was found between social characteristics (marital and employment status) and central obesity [14]. Theoretically, marriage will cause couples to share sedentary living habits, obesogenic diets, or lose motivation to maintain body shape [31]. Meanwhile, employment status is closely associated with the energy expenditure level, thus influences the energy balance [17]. The insignificant relationship found in this study might be due to homogenous participants’ characteristics of both risk factors.
Central obesity was more likely to be experienced by participants with low education levels and poor nutrition knowledge, although the relationship was insignificant. Good education and knowledge are associated with better application of healthy lifestyles [16]. In contrast, Zhou et al. stated that individuals with good education and knowledge do not necessarily have complete nor applicable information so that they can prevent and overcome specific nutritional problems [32]. Even though there were more participants with high education level and good nutritional knowledge, poor diet habit (e.g. excessive energy and macronutrient consumption, insufficient dietary fiber consumption) was more common. In conclusion, knowledge may not be manifested in behavior [17].
Total blood cholesterol level was found to be the dominant factor of central obesity. The odds of central obesity were 4.210 times higher for the ‘hypercholesterolemia’ group compared to the ‘normal total blood cholesterol level’ group. The median total blood cholesterol levels were high, although those might be influenced by the testing method (non-fasting condition and usage of electrode-based biosensor testing device can lead to higher measurement results). While the relationship between hypercholesterolemia and central obesity has not been widely proven in epidemiological studies, various clinical studies have found the mechanism that links both concepts. Individuals with hypercholesterolemia (with normal or high BMI) have a common phenotypic characteristic, which is increased WC [32]. The mechanisms of central obesity in hypercholesterolemia patients involved various metabolic mechanisms, namely oxidative stress, adipose tissue dysfunction, and hormonal dysregulation, hence increase de novo lipogenesis in intra-abdominal organs [18].
This study did not find a significant relationship between psychological stress level and central obesity, similar to findings in a previous study in Bogor by Sudikno et al. [12]. Previous research in different areas have shown that the prevalence of central obesity was higher among individuals with psychological stress [14, 17]. People with chronic stress conditions experienced long-term low levels of systemic inflammation and an increase in appetite, thus increase the tendency of being centrally obese [17]. Further analysis showed that the level of energy and macronutrient intake, physical activity, and smoking habits between ‘low’ and ‘moderate to high’ stress level groups were not significantly different.
In general, participants had a poor dietary habit, expressed by excessive dietary intake, particularly from fat sources. The average energy and macronutrients consumed were exceeding the requirements (energy 150.58±70.75%EER, protein 129.62±58.94%NR, fat 251.14±114.62%NR, and carbohydrate 129.78±46.52%NR). The amount of energy derived from fat was excessive (avg. 37.7±9.47%vs. recommended 10–25%TEI), sufficient from protein (avg. 12.4±2.14%vs. recommended 10–15%TEI), and deficient from carbohydrate (avg. 48.4±9.76%vs. recommended 60–75%TEI). The excessive energy and macronutrient intake were suspected to be associated with high consumption of animal protein, simple carbohydrates, and fried foods. Unfortunately, this hypothesis cannot be confirmed since the data received from SQFFQ was already converted to kcal or grams of nutrients.
A significant relationship between central obesity and excessive energy, protein, fat, and carbohydrate intake was found (OR = 6.188; 4.929; 9.158; dan 2.970). Positive energy balance occurs when energy intake is greater than energy expenditure. The excess energy derived from protein, fat, and carbohydrate catabolism will be stored in subcutaneous adipose tissue first, and when it reaches the maximum storage capacity, fat will accumulate on the surface of intra-abdominal organs [33]. Continuous excessive energy intake will induce leptin resistance, so that satiety no longer reduces appetite [35]. As a result, a person cannot control the amount of food consumed even though already obese.
The relationship between dietary fiber intake and central obesity was insignificant. Similar to the current study, a previous study did not find a significant relationship between fruits and vegetable consumption with central obesity [12]. The median dietary fiber consumed by participants was 16.00±10.41 grams/day. This finding suggested that the adult population in the study area is having difficulty meeting the daily fiber consumption of 25 grams/day (consist of 3-4 servings of vegetables and 2-3 servings of fruits daily), as recommended by the Ministry of Health of Indonesia.
The prevalence of central obesity was higher among participants with ‘>300 mg/day’ compared to ‘≤300 mg/day’ cholesterol intake, although the difference was not significant. While a high cholesterol diet may contribute to hypercholesterolemia [21], further analysis of this study found no significant relationship between cholesterol intake and hypercholesterolemia. The body will eliminate excessive cholesterol metabolites by excreting them as free cholesterol in bile and bile salts [36]. Thus, occasional excessive cholesterol consumption will not cause hypercholesterolemia, and it is not directly related to central obesity.
The current study found that physical activity and smoking habits were not significantly related to central obesity. Theoretically, unhealthy living behaviors, such as low physical activity level and smoking habit, increase a person’s risk of developing central obesity [15, 22]. The majority of participants had an active lifestyle and did not smoke. It is important to note that the use of the retrospective method in GPAQ might involve some bias that lead to over-interpreted physical activity levels. The homogeneity found in physical activity and smoking habit data might be the cause of the insignificant relationship found between those risk factors and central obesity.
Conclusion
This study proved an increasing trend of central obesity in young adults. Unhealthy diet habits, as expressed by excessive fat intake and low dietary fiber consumption, were the common findings among participants. Female sex, hypercholesterolemia, as well as excessive energy and macronutrient intake, were significantly related to central obesity in adult patients at Bojong Gede Public Health Center, Bogor Regency, West Java Province, Indonesia. Hypercholesterolemia was the dominant factor of central obesity after adjusted for confounding variables.
The prevention and control efforts for central obesity should be focused on health promotion aimed at improving the quality and quantity of diets. Large-scale screening for central obesity and hypercholesterolemia can be integrated. The program should not only target the late-adulthood but early-adulthood age group as well.
Abbreviations
BMI: Body Mass Index; CI: Confidence Intervals; COVID-19: Corona Virus Disease 2019; EER: Estimated Energy Requirement; GCU: General Check-Up; GPAQ: Global Physical Activity Questionnaire; HDL: High-Density Lipoprotein; LDL: Low-Density Lipoprotein; MET: Metabolic Equivalent of Tasks; NR: Nutritional Requirement; OR: Odds Ratio; PSS: Perceived Stress Scale; SPSS: Statistical Package for the Social Sciences; SQFFQ: Semi-Quantitative Food Frequency Questionnaire; TEI: Total Energy Intake; WC: Waist Circumference.
Ethics Approval and Consent to Participate
The study protocol was approved by The Ethics Commission of Faculty of Public Health Universitas Indonesia (No. 61/UN2.F10.D11/PPM.00.02/2020). Legitimate data usage permission was obtained from the primary study researcher. Written informed consent was obtained from all subjects.
Conflict of Interest
The authors declare that they have no competing interests.
Availability of data and materials
The datasets analyzed during the current study are available from the corresponding author on reasonable request.
Author’s Contribution
Benedicta Natalia Latif wrote the draft paper, analyzed the data, and interpreted the results. Ratu Ayu Dewi Sartika and Fani Widiartha designed the study and collected the data. Benedicta Natalia Latif and Ratu Ayu Dewi Sartika revised the manuscript. All authors have approved the final article.
Footnotes
Acknowledgments
The authors would like to thank Fani Widiartha for permitting the use of her master thesis data for this research. Data were collected in 2017 on self-funding.