Body mass index and weight status misperception among a sample of college students in Yogyakarta , Indonesia

Background: Weight status perception associates with objective weight status and is important in the management of weight control. To date, perception of weight status among Indonesian youths has not been reported. Objectives: This study aimed to examine the association between body mass index and weight status perception in a sample of college students in Yogyakarta Province. Materials and Methods: A sample of 209 boys and 269 girls of college students in Yogyakarta Province were measured for their stature and body weight. Body mass index was calculated (BMI). Data of demographic, exercise, and diet were collected. Weight status perception was based on participant responses to a question regarding how they classified their own body size as underweight, normal, overweight, or obese. Ordinal regression analysis was performed to evaluate factors associated with weight status misperception among boys and girls. Results: Overall, 43.5% of boys and 37.5% of girls misclassified their own weight status by actual BMI. Of particular note, 75.9% of obese boys and 78.6% of obese girls underestimated their weight status as overweight or normal weight. Whereas, 9.1% and 23.4% of normal weight boys and girls respectively, overestimated their weight status. Ordinal regression analysis revealed that, weight status misperception from others was significantly contributed (p<0.01) to misperception of weight status among boys and girls with OR of 10.31 and 8.13 respectively. Diet practicing was significantly correlated with weight status misperception in boys (p<0.05) with an OR of 19.57. Conclusions: Weight status misperception was prevalent among normal weight and obese students. Obese students of both gender and normal weight boys tended to underestimate their weight status, whereas normal weight girls were likely to overestimate their weight status. Bangladesh Journal of Medical Science Vol.16(2) 2017 p.225-232


Introduction
Overweight and obesity prevalence globally increase at an alarming level.Data from the Ministry of Health, Republik Indonesia 1,2 showed that the prevalence of overweight and obesity increased from 18.9% in 2007 to 21.3% in 2013.It is widely known that overweight and obesity lead to some major diseases.World Health Organization reported that increasing degrees of overweight, as measured by body mass index (BMI) increases co-morbidity risks including coronary heart disease (CHD), stroke, type 2 diabetes, and some cancers, and mortality rates 3 .Accordingly, public health action in the managementof prevention and treatment efforts for obesity is straight away essential.
In attempt to design appropriate health advice to intervene obesity, understanding factors associated with obesity is necessary.Among the factors is body image, and body size perception is among body image interests, which has been studied in relation to obesity [4][5][6][7][8] .A number of studies have examined levels of overestimation and underestimation of body size in youths compared to their actual body size obtained from e.g.BMI.Reports indicated that a large proportion of boys and girls has misperception about their weight status. 5 Boys were likely than girls to classify their weight status lower than their weight status determined by actual BMI, hence, considered as underestimation of weight status.Whereas girls were more likely to classify their weight status higher than their actual weight status measured with BMI which considered as overestimation of weight status 8 .Some studies indicated that a substantial proportion of obese youths misperceived their weight status as overweight or normal weight instead their clinical obese category 6,8,9 .However, findings were not consistent that overweight youths rather than obese ones were more likely to underestimate their weight status 7 .Moreover, weight perception of being in the normal weight rather than weight misperception was associated with better health related quality of life among youths 10 .Perception of being overweight and underweight rather than weight status misperception were significant factors associated with psychological distress 11,12 .Meanwhile, perception of overweight is an important determinant of nutritional habits and weight management in youths, overweight youths who did not perceive themselves as such were unlikely to engage in weight control practice 6 .
Several factors were thought to influence weight misperception among youths.A study in Mexican youths found that over estimating was associated with female gender, younger age, lower level of parent education, and more hours of daily TV viewing, while, underestimating was associated with male gender and older age 8 .Obesity exposure increased acceptance of obesity in obese male gender 13 .Exposure to overweight or obese environment in which people they see on a daily basis predicted misperception of weight status 9 .Yet, no studies have examined weight status misperception among Indonesian youths and factors that might predict weightmisperception.Therefore, the current study aims to investigate the association between perceived weight status and actual BMI in a sample of Indonesian college students, particularly those who were living in Yogyakarta Province.Demographic factors, information about exercise and dieting, as well as weight perception from others were evaluated with regard to their contribution to weight misperception among Indonesian youths.

Participants
A number of 209 male and 269 female students of Universitas Gadjah Mada (UGM) and Universitas Teknologi Yogyakarta (UTY) living in Yogyakarta Province, aged 17-25 years participated in the present study.Participants were studentsof Grade 1 st to 4 th at the Faculty of Medicine and Faculty of Cultural Sciences UGM, and Faculty of Information Technology and Business UTY.Majority of the participants were Javanese (86.3%).Participants having physical disabilities or in pregnancy for female students were excluded from the study.
A structured questionnaire was administered to participantscovering gender, background information (birth and place of date, ethnicity, school grade, parents' education,parents' income,diet experiences, engagement in regular physical activity, and perception of their weight status from other people (their peers or family).Parents' education level(father and mother) was categorized into high school or lower education levels and college graduation.Parents' income level (father and mother) was determined by less than or equal to IDR 2.5 million (approximately USD 200) monthly or more that value.Diet practicing and physical activity was observed by answering "yes or no" questions whether engage or not in those activities.Weight status from other people was the way of other people classified participants' weight statusi.e.underweight, normal, overweight, or obese.Participants signed and returned an informed consent form.The study protocols have been approved by the Medical and Health Research Ethics Committee of Faculty of Medicine UGM.

Measures
Actual weight status.Body weight and stature were measured using the standard protocol of the International Society for the Advancement of Kinanthropometry (ISAK).Stature was taken using an anthropometric set (GPM, Swiss, Ltd.) to the nearest 0.1 cm.Body weight was measured with the participant wearing light clothing with a Seca weight scale (Seca 803, Seca Deutschland) to the nearest 0.1 kg.BMI was calculated as kg/m 2 and was used as the measure of actual weight status.Participants were categorized into underweight (<18.5 kg/m 2 ), normal (18.5-24.9kg/m 2 ), and overweight (25.0-26.9kg/ m 2 ), and obese (≥27.00 kg/m 2 ).The classification is adopted from World Health Organizationadjustedfor Indonesians 14 .
Perceived Weight status.Participants reported perception of their own weight status by answeringa question "How do you classify your body at this moment?"Answers were chosen among four options: underweight, about normal weight, overweight, and obese.Participants were classified into three categories: about the right weight, overestimated, and underestimated weight status.Participants whose actual BMI category was greater than their perceived weight status were consideredas "underestimated" weight status perception and vice versa.While, those who accurately perceived their selves as their actual BMI category were recognized as having "about the right weight".

Statistical Analysis
Differences of characteristics between boys and girls were determined using t-student test for continuum scale and chi-square test for ordinal scale.Ordered regression analysis was performed to evaluate factors contributed to weight status misperception among boys and girls.A p-value of <0.05 was considered significant.Magnitude of association was expressed as an odd ratio with a 95% confidence interval.All statistical analyses were done using the SPSS program (version 20.0, SPSS Inc., 2011, Chicago, IL).

Results
Characteristics of participants are summarized in Table 1.Boys were significantly taller, heavier, and greater BMI (p<0.01)than girls.Differenceswere also existed (p<0.01) in the distribution of BMI category, school grade, and engagement in regular physical activity.Obesity prevalence (including overweight and obese) was higherin boys (21.6% vs 8.2%), however, self-weight perception for obesity was almost similar in both genders.  2 displays the prevalence of weight status misperception among actual weight status classified according to BMI.Totally, 43.5% of boys and 37.5% of girls misclassified their weight status relative to these medical standards.A substantial percentage of boys classified themselves as having lower weight status than their actual weight category (32.5%), by contrast, girls showed a greater percentage of those who overjudged their weight status as heavier than their actual weight category (27.1%).However, with respect to those who were determined as obese according to actual BMI, most of them perceived their selves as overweight or even normal weight instead both in boys (75.9%) and girls (78.6%).Contributions of some factors in the weight status misperceptions among boys and girls are presented in Table 3.This tables describes the results of ordinal regression analysis with weight misperception as independent factors and parents' education and income, diet practicing, physical activity, and weight status perception from others as dependent factors.Among those factors, weight status perception by others was a significant factor in both genders and diet practicing in boys.In boys, the odds of right perception by others considering misperception by self was 0.097 (95%

Discussion
This study examined the association between perceived weight status and actual BMI among college students in Yogyakarta Province, Indonesia.Findings indicate that a substantial number of boys and girls misperceived their weight status i.e. 33.5% and 37.5% in boys and girls, respectively.Boys showed a greater prevalence of underestimation, while, girls were more likely to overestimated their weight status.Of particular note, among obese students by medical category, only about one forth in boys and one fifth in girls were correctly identified their weight status, while the rest were underestimated themselves as lower weight status.
Ordinal regression analysis reveals that misperception by others was 10.31and 8.13 times that of right perception by others considering misperception by selves in boys and girls respectively (p<0.001).Diet practising was significantly associated with weight status misperception in boys only with an odds of 19.57times (p<0.05)greater than those who did not involve in diet.Whereas, the other factors including parent's education and income, and engagement in regular physical activity were not significantly contributed to weight status misperception.Our study was consistent with previous research that higher incidence of girls of college students overestimated their weight status compared with boys (23.9% vs. 13.9%).By contrast, higher prevalence of boys was underestimated their weight status 8 , The current study also indicated that agreement between actual BMI and weight status perception was high among underweight, normal, and overweight (62.4% to 91.7%), but poor among obese in both genders (23.3% males and 14.3% females).Overestimation of weight status existed among underweight and normal BMI categories in our samples with normal weight girls showing higher rate of overestimation than boys do.The mismatch of weight status perception and actual weight status was also reported among college students in previous studies in Nigeria 15 and Mexico 8 .McCabe and colleagues 16 suggested that perceptual accuracy of body image might highly associated with biopsychosocial influence.That sociocultural norms and media influences associated with overestimation of weight status has been acknowledged.Among girls, overestimation on weight status in girls was strongly influenced by high levels of depression, media, and peer influences to be thinner.Whereas boys experienced overestimation of body size were predicted by high BMI, media influences to lose weight and increase muscle, and peer influences to increase muscle 16 .Younger age, female gender, lower level of parent education, and daily TV viewing were also predicted overestimation 8 .Media influence and TV viewing were not observed in the current study, however, college students were highly expose with internet connectivity due to high load of assessment and learning material preparations which may potentially expose the students to media influence.It is required further study that a longer time of exposure to the internet connectivity was also suggested to influenceweight status perceptionthrough sedentary life style ora standard of beauty that exaggerated slimness fashionable.Engagement in physical activity and diet practising were not significantly predicted overestimation of weight status in the present study, especially in girls.Low prevalence of girls who engage in those activities (7.5% in regular exercise and 14.1% in diet practicing) might influence these results.
An important finding of this studyis the high prevalence of obese students (>75%) who misperceived their weight status as having lower weight category.It was suggested that underestimation of overweight was correlated with BMI, intense physical activity, knowledge of a healthy weight range, and body comparison 17 .More recent studies also indicate that exposure to heavier body weight increased the acceptance of obesity, particularly in male gender.Exposure to obese people resulted in an obese male being judged more positively than after exposure to healthy weight 13 .There was also an evidence of a generational shift in social norms related to body weight resulted in people may be less likely to intend weight loss.This was shown on male gender that exhibited a highly decrease in feeling overweight than normal weight 18 .Congruently, in the current study, weight status perception from peers showed significant predictor of weight status misperception in both gender.This emphasized the importance of public health action to provide knowledge concerning the right weight status to overall youth population.The program should aims to educate youths to improve body image, besides emphasizing the health risks associated with excess body weight.
Nevertheless, several limitations should be kept in mind when interpreting these data.First, we cannot suggest that our data are representative of the overall Indonesian students, or Indonesian youths with similar age range, since association between BMI and weight status perception differ across ethnics [19][20][21] .
Our samples are college students living in Yogyakarta Province which is only one among 33 provinces in Indonesia.However, due to a considerable number of ethnicities, most of the students in our samples are taken from Javanese population which represents the largest ethnicity in Indonesia.Second, our data have low prevalence of students with overweight and obese categories, especially female which is less than one third of the national prevalence of obesity 2 .This might be influenced by the small range of age (17-24 years) in the early adult.Increased BMI and age has been reported associated with body image 4,22 .Third, we did not investigate several factors that might associated with perception of weight status, such as psychological distress, media influence, and weight control behaviours.Evaluation on these factors might help formulating issues need to be addressed to prevent misperception of weight status and further risks.Lastly, it may also be a limitation of this study due to the less variability or small sample size.A nationally representative of sample size and Indonesian populations is substantial in advanced studies.
Despite those limitations, our study demonstrates essential evidence of misperception of weight status among Indonesian youths whose age period is susceptible to engage in dangerous behaviours leading to health risks related to weight status.A higher structured design and more comprehensive study involves ethnic diversity is important to provide nationally representative data.Future studies should also include broader biopsychosocial assessment to identify factors associated with weight status misperception among Indonesian youths.Understanding factors that contribute to misperception of weight status in Indonesian youths from diverse ethnicities and sociocultural background is worthwhile to appropriate design intervention strategies to promote an understanding of healthy weight.

Table 4 . Ordinal logistic regression analysis of factors contributed to weight misperception in boys and girls
CI, 0.046 to 0.204) times that of misperception by others (p<0.001).Or, inversely, misperception by others was 10.31 times that of right perception by others considering misperception by self.The odds ofdiet experience considering misperception by self in boys was 4.766 (95% CI, 1.161 to 19.572) times that of not diet experience(p < 0.05).Whereas, in girls, the odds of right perception by others considering misperception by self was 0.123 (95% CI, 0.065 to 0.234) times that of misperception by others(p<0.001).Or, inversely, misperception by others was 8.130 times that of right perception by others considering misperception by self.