The Prevalence of Underweight, Overweight/Obesity and Their Related Lifestyle Factors in Indonesia, 2014–2015

Objective To quantify the prevalence of underweight and overweight or obesity and its related factors (socio-demographic, health behavior, health status) in a national adult population in Indonesia. Material and Methods In a national cross-sectional population-based survey in 2014–15 in Indonesia, 29509 adults (median age 41.0 years, Inter Quartile Range = 22.0, age range of 18–103 years) completed questionnaires and anthropometric measurements. Multinomial logistic regression modelling was used to determine the association between socio-demographic, health behavior and health status factors and underweight and overweight or obesity. Results Of total sample (n = 29509), 11.2% measured underweight (13.5% among men and 9.1% among women) (<18.5 kg/m2), 39.8% normal weight (48.1% among men and 32.0% among women) and 49.0% had overweight or obesity (≥23 kg/m2) (38.3% among men and 58.9% among women); 24.6% of the overall sample had class I obesity (25–29.9 kg/m2), and 8.5% had class II obesity (30 or more kg/m2). Among different age groups, underweight was the highest among 18–29 year-olds (20.0%) and those 70 years and older (29.8%), while overweight or obesity was the highest in the age group 30 to 59 years (more than 53%). In adjusted multinomial logistic regression, having less education, living in rural areas and not having chronic conditions were associated with underweight status. While better education, higher economic status, urban residency, dietary behavior (infrequent meals, frequent meat, fried snacks and fast food consumption), physical inactivity, not using tobacco, having chronic conditions (diabetes, hypertension, hypercholesterol), and better perceived health and happiness status were associated with overweight or obesity. Conclusions A dual burden of both adult underweight and having overweight or obesity was found in Indonesia. Sociodemographic, health risk behavior and health status risk factors were identified, which can guide public health interventions to address both these conditions.


Introduction
Globally, from 1975 to 2014, the prevalence of underweight decreased from 13.8% to 8.8% in adult men and from 14.6% to 9.7% in adult women, and the prevalence of obesity increased from 3.2% to 10.8% in adult men, and from 6.4% to 14.9% in adult women [1]. In Indonesia in 2013, the national adult (18 years and older) prevalence rate of underweight was 11.1% and the prevalence of overweight or obesity (≥25 kg m 2 ) was 30.4% [2,3]. The prevalence of underweight and overweight among girls (15-18 years) in rural areas was 25% and 10%, respectively, and in urban areas 28% and 10%, respectively, in Indonesia [4]. Similar rates of underweight and overweight or obesity have been reported in countries of the Southeast Asian region. In Bangladesh (35 years and older), 30.4% were underweight and 23.5% had overweight or obesity (≥ 23 kg/m 2 ) [5]; in Malaysia (18 years and above), 51.2% had overweight or obesity (≥ 25 kg/m 2 ) [6]; in Thailand, 40.9% had overweight or obesity (≥23 kg/m²) [7]; and in Vietnam in 2005, 20.9% were underweight and 16.3% had overweight or obesity (≥23 kg/m²) [8]. As globally, a decrease in the prevalence of underweight and increase of overweight or obesity over the past 20 years have been reported in Indonesia [9] and other countries in the Southeast Asian region such as Vietnam [8] and Thailand [10].
Undernutrition in adulthood can lead to increased morbidity and mortality and other adverse outcomes [11]. Obesity is a major risk factor for a number of non-communicable diseases, such as cardiovascular disease, hypertension, stroke, diabetes mellitus, and specific forms of cancer leading to increased morbidity and premature mortality [12]. Factors related to adult underweight may include sociodemographic variables, such as being female [8,13,14], early adulthood (15-24 years) [14], older adults [5,14], having lower education [5,[13][14][15], poorer economic background [5,14,15], not working [5], being married [15], being a non-Christian [15] and residing in rural areas [8]. For example in India, among "rural young (15-24 years) females from more educated villages had a higher likelihood of underweight relative to those in less educated villages; but for rural mature (>24 years) females the opposite was the case" [15]. Moreover, food insecurity, inadequate food intake, diets low in diversity and with insufficient nutrient density [4] and fear of being fat [16] may be associated with underweight.
There is a lack of more recent national data on the prevalence of underweight and overweight and obesity and its sociodemographic, behavioral and physical and mental health risk factors in Indonesia.
It is important to understand factors driving the dual underweight and obesity burden in Indonesia.

Method
Data were analysed from the "Indonesia Family Life Survey (IFLS-5)", a continuing demographic and health survey that began in 1993 and had since four rounds of data collection, with the fifth wave (IFLS-5) fielded in late 2014 and completed in 2015 [35]. The community surveys collected data on household, individual, and community level using a multistage stratified sampling [35]. The sampling frame of the first survey in 1993 was based on households from 321 enumeration areas (EAs) (20 households were randomly selected from each urban EA, and 30 households were selected from each rural EA) in 13 out of 27 Indonesian provinces that were selected representing 83% of the Indonesian population in 1993, more details in [35]. At household level, several randomly selected household members were asked to provide detailed individual information. Being a member of a selected household (15 years and above) was defined as members that reside in the same dwelling and share food from the same cooking pot [35]. In the IFLS-5, 16,204 households and 29509 (excluding 617 who were pregnant) 18 years and older individuals were interviewed with complete anthropometric measurements.
In the IFLS-5 the household response rate was 90.5% [35]. Although the survey is longitudinal, we restricted our analysis to the IFLS-5 cross-sectional survey for persons 18 years and older, being the most recent national survey available assessing anthropometric measures.
Socio-demographic factor questions included age, sex, marital status, education, work status, religion, residential status, subjective socioeconomic background, and country region. Subjective economic status was assessed with the question "Please imagine a six-step ladder where on the bottom (the first step), stand the poorest people, and on the highest step (the sixth step), stand the richest people.
Childhood hunger was assessed with the question "Did you experience hunger in your childhood (from birth to 15 years)?" [35] (Coded as yes).
Fruit consumption was assessed with three questions on, "How many days in the past week, (1) Responses were coded as 1 = 3 times/day and 0 = 2 times/day or less.
Physical activity was assessed with a modified version of the "International Physical Activity Questionnaire (IPAQ) short version, for the last 7 days (IPAQ-S7S)". We used the instructions given in the IPAQ manual [37], and categorized physical activity according to the official IPAQ scoring protocol [38] as low, moderate and high.
Tobacco use was assessed with two questions: (1) "Have you ever chewed tobacco, smoked a pipe, smoked self-enrolled cigarettes, or smoked cigarettes/cigars?" (Yes, No) (2) "Do you still have the habit or have you totally quit?" (Still have, Quit) [35]. Responses were grouped into never, quitters and current tobacco users.
Self-reported health status was measured with the question, "In general, how is our health?" Response options were 1 = Very healthy, 2, Somewhat healthy, 3 = Somewhat unhealthy, and 4 = Unhealthy [35].
Diabetes or high blood sugar, High blood pressure, High cholesterol (total or LDL) were assessed with the question, "Has a doctor/paramedic/nurse/midwife ever told you that you had…?" (Yes, No) [35].

Blood pressure (BP) measurements and classification
Three consecutive measurements of systolic and diastolic blood pressure (BP) were recorded with an Omron meter, HEM-7203, by regular trained interviewers on household members 15 years and older at home in a seated position [35]. Information was collected on awareness and treatment of hypertension. Average blood pressure was calculated arithmetically for the three measurements of each systolic and diastolic blood pressure. Blood pressure classification was done using JNC 7 algorithm [39]. Hypertension was defined as SBP ≥140 mm Hg and/or DBP ≥90 mm Hg and/or current use of antihypertensive medication. Normotension was defined as BP values <120/80 mm Hg in individuals who were not taking antihypertensive medication [39].
Sleep disturbance was assessed with five items from the "Patient-Reported Outcomes Measurement Information System (PROMIS)" sleep disturbance measure [40]. A sample item was, "I had difficulty falling a asleep." Response options ranged from 1=not at all to 5= very much.
(Cronbach's alpha = 0.68 in this study). Sleep disturbance was defined as a score of four or five on the averaged mean items.
The Centres for Epidemiologic Studies Depression Scale (CES-D: 10 items) was used to assess depressive symptoms, and scores 10 or more were classified as having depressive symptoms [41] (Cronbach's alpha = 0.71 in this study).
Happiness was assessed with the question, "Taken all things together how would you say things are these days -would you say you were very happy, happy, unhappy or very unhappy?" [35]

Data Analysis
Descriptive statistics were used to describe the variables. Multinomial logistic regression analysis was computed to calculate the adjusted relative risk ratios (ARRR) with 95% confidence interval (CI) to determine the associations between socio-demographic, health risk behavior, physical and mental health risk status variables and underweight and overweight or obesity. Potential multi-collinearity between variables was assessed with variance inflation factors, none of which exceeded critical value of 4. P < 0.05 was considered significant. Missing values were excluded from the analysis.
"Cross-section analysis weights were applied to correct both for sample attrition from 1993 to 2014, and then to correct for the fact that the IFLS1 sample design included over-sampling in urban areas and off Java. The cross-section weights are matched to the 2014 Indonesian population, again in the 13 IFLS provinces, in order to make the attrition-adjusted IFLS sample representative of the 2014 Indonesian population in those provinces." [35] Both the 95% confidence intervals and P-values were adjusted considering the survey design of the study. All analyses were done with STATA software version 13.0 (Stata Corporation, College Station, TX, USA).

Sample Characteristics
The total sample included 29509 adults (median age 41 The prevalence of underweight was significantly higher among male young adults (20.0%) than female 18-29 year-olds (13.2%), and in the age groups 30-49 years, while the prevalence of underweight in age groups 50 years and above was still higher among men than women but not significantly. Among the sexes, the prevalence of overweight or obesity was significantly higher in women than men in every age group (see Table 1). depression symptoms, and 11.8% rated themselves as being "very happy" (see Table 2).    Table 3).

Discussion
The results of this national study demonstrate the co-existence of dual burden of underweight   [2,3]. Largely similar rates of underweight and overweight or obesity have been reported in countries of the Southeast Asian region such as Bangladesh [5], Malaysia [6], Thailand [7] and Vietnam [8]. Indonesia. This trend is consistent with a global decrease in the prevalence of underweight and increase of overweight or obesity over the past 20 years, which has also been observed in countries of the Southeast Asian region such as Vietnam [8] and Thailand [10].
The study found that the prevalence of underweight was the highest among 18 to 29 year-olds (16.8% overall, 20.0% in males and 13.2% in females) and those 70 years and older (29.8%). Previous studies also confirm the higher prevalence of underweight among young adults [14] and older adults [5,14]. In this study the prevalence of underweight was in bivariate analysis significantly higher among men than women, while previous studies found an opposite result [8,13,14]. It is not clear why the prevalence of underweight is higher in men than women, partially also explained by a lack of data reporting undernutrition in adult men [5]. Reasons for the high prevalence of underweight during early adulthood may be related to food insecurity [4] and fear of being fat [16]. Some studies report an increase of an underweight body ideal and in eating disorders in Southeast Asia, including Indonesia [42].
In a small study among university students in Indonesia 14.5% of male and 5.1% of female students reported disordered eating attitudes.
In agreement with previous studies [5,8,[13][14][15], this study found an association between lower education, not working and living in rural areas with underweight. Unlike some previous studies [4,5,14,15], this study did not find an association between poorer economic background, the experience of childhood hunger and underweight. The association between participants attending school, retired, sick or disabled and underweight may be related to the higher prevalence of underweight among adolescents and older adults. Compared to persons living in Sumatra and Java, individuals living in major island groups (Bali, South Kalimantan, South Sulawesi and West Nusa Tenggara) had a higher prevalence of underweight. Hanandita and Tampubolon [14] found that undernutrition is spatially clustered within islands of Indonesia.
Regarding overweight or obesity, in consistence with previous studies [5,8,14,15], this study found that being female, being middle aged, having higher education, higher economic status and urban residence were associated with having overweight or obesity. Compared to persons living in Sumatra, individuals living in Java major island groups had a lower prevalence of overweight or obesity. Hanandita and Tampubolon [14] found that over-nutrition is spatially clustered within islands of Indonesia. Further, specific dietary behavior (infrequent meals, frequent meat, fried snacks and fast food consumption) was, as previously found [18][19][20][21][22]25,26], associated with having overweight or obesity. It is likely that eating frequency impacts obesity through a combination of different mechanisms affecting a decrease in hunger, an increase of satiety responses and a reduction binge eating. Contrary to some studies [19,23], this study found that inadequate fruit consumption and frequent sweet snacks consumption were negatively associated with having overweight or obesity.
These findings largely confirm a shift in dietary behavior towards increased consumption of calorie-dense foods containing refined carbohydrates, fats, red meats, and low fiber in Indonesia.
In line with previous studies [6,18,19,27,28], this study also found a negative association between physical activity, smoking and having overweight or obesity. In this study, however, the association between physical inactivity and overweight or obesity was only observed in men and not in women (in sex-stratified analysis, not shown). Similar results were found in the Malaysian national study [6]. A possible explanation for this result could be that a greater proportion of men than women engage in high physical activity producing an effect large enough to reduce obesity [6].
The study findings further suggest, as previously found [3,6,29], that cardiometabolic comorbities, including diabetes, hypertension and dyslipidaemia, were associated with having overweight or obesity. Contrary to a previous study on self-rated health status [28], and in agreement with a previous study on happiness [34], this study found an association between very healthy self-rated health status, very happy happiness status and having overweight or obesity. Moreover, unlike previous studies on depression [30,31], this study found a negative association between depression and having overweight or obesity. Some previous study [33] found an association between poor sleep quality and having obesity, while this study did not find such an association.
Perhaps, overweight or obesity is further shaped by the ideal body image of fatness as a symbol of nurturance and affluence [14,34]. Sohn [34] proposes that the positive relationship between having obesity and happiness may be "explained by improved socioeconomic status and health enjoyed by the obese". In case Indonesians perceive overweight or obesity status as a signal of better health and greater happiness, it will be more difficult to target obesity in intervention programmes.

Study limitations
The study was a cross-sectional study and the temporal relationships between sociodemographic factors, health status and health risk behaviors and underweight and overweight or obesity cannot be established in such studies; further longitudinal studies are needed. Apart from anthropometric and blood pressure measurements, a limitation of the study was that all the other information, including physical activity, was collected based on self-reporting.

Conclusion
The paper found in a cross-sectional study a high prevalence of both underweight and overweight or obesity among 18 years and older individuals in 2014-2015 in Indonesia, with underweight decreasing and overweight or obesity increasing in the past 20 years, as assessed in previous other studies. Sociodemographic, health risk behavior and health status risk factors were identified which can guide much needed public health interventions to address both these conditions.