Introduction

A slim figure is the socially acceptable body size among female adolescents and young women, especially in industrialized countries. This is being perpetuated by role models in the electronic media and magazines as well as by peer pressure (Taylor et al, 1998; Borzekowski et al, 2000; Tiggemann et al, 2000). Consequently, there is widespread preoccupation with weight control to achieve such a body size (Tiggemann et al, 2000; Abbott et al, 2000). This may be beneficial as overweight adolescents have a higher risk of becoming overweight adults (Stunkard & Burt, 1967). In addition, being overweight or obese during adolescence, even if the individuals are of normal weight as adults, can result in increased morbidity and mortality from a wide range of diseases (Must et al, 1992). However, the obsession with thinness is not entirely good. Young adults, adolescents and even pre-teens tend to engage in a variety of behaviours in order to lose weight, some of which are very dangerous, in order to achieve what they perceive to be the ideal body size (Moore, 1988; Striegel-Moore, 2001; Ricciardelli & McCabe, 2001).

The body image of adolescents has been addressed in many research studies. This is not surprising, given its association with eating disorders and other psychiatric problems such as depression (Killen et al, 1996; Taylor et al, 1998; Stice et al, 2000). Much of the focus has been on white females, probably because of the belief that there are cultural differences with respect to body image and the ideal body size. For example, the results of some studies indicate that black females are more likely to be satisfied with their size and had less desire for a slim figure than white females in the USA (Strauss, 1999; Miller et al., 2000). However, other studies found no difference between black and white adolescents (Schreiber et al, 1996; Striegel-Moore, 2001; Fitzgibbon et al, 2000). There is some evidence that part of the reason for the apparent inconsistency in the findings may be related to social class differences. Indeed, the results of one study indicated that more black females of higher social classes had concerns with being overweight than their counterparts from lower social classes (Robinson et al, 2001). Other ethnic groups such as Hispanics and Asians living in the USA have also been found to strive for a slim figure (Robinson et al, 1996); however it is likely that culture has some influence on adolescents' body image even though its role is not fully understood.

The body image of adolescents in developing countries such as those in the Caribbean has not been the focus of much research and only one published report has been identified. In that study, the perceptions of female, Jamaican adolescents were examined and the authors concluded that the girls had an accurate perception of their size (Smith & Gogswell, 1994). Unfortunately, they did not examine satisfaction with body size. We conducted the present study in Trinidad, which is not only in a state of economic transition but has a multi-ethnic population. Our study facilitated a comparison of the results with those of industrialized countries as well as the determination of whether there were differences in the body image among the various ethnic groups. We included both males and females in the sample because, although body image problems are greater among females (Kiefer et al, 2000; Houn & Brown, 1986), males can also be affected (Robinson et al, 2001; Striegel-Moore, 2001).

Methods

The study design was a cross-sectional survey. It included students in their fourth year of secondary schools (Fourth Form) in Trinidad. The study was approved by the Ethics Committee of the Faculty of Medical Sciences, University of the West Indies and the Ministry of Education, Trinidad and Tobago.

The sample was selected using stratified random sampling. The Ministry of Education in Trinidad has divided secondary schools into zones, according to their geographical location. The schools were stratified by zone and 16 of the 73 secondary schools were randomly selected. The number of schools selected from each zone was proportional to its student enrollment. Fourth-form students were then chosen from each selected school using simple random sampling.

A self-administered questionnaire was used to measure demographic variables, perceived body size, the most attractive body size (of the opposite sex), whether they were satisfied with their current size, whether they would like to change it, and the body sizes that they associated with health, wealth and happiness. The latter were included because Caribbean people have traditionally associated fatness with wealth and happiness (Hoyos & Clarke, 1987) and it was not clear whether such beliefs were present among adolescents. Silhouettes were used to assist in the determination of perceived and most attractive body sizes as well as in the identification of the sizes associated with health, wealth and happiness. Four male and four female silhouettes were used corresponding to thin, normal, overweight and obese body sizes (Figure 1). To determine actual body size, weight and height were measured using standard techniques (WHO, 1995).

Figure 1.

The silhouettes used in the perception of body size (1=thin, 2=normal, 3=overweight, 4=obese).

Full figure and legend (22K)

The weights and heights were used to calculate body mass index (BMI), ie weight (kg)/height (m)². Children were then classified as being thin (<5th percentile), normal (5th–85th percentile) or overweight/obese (>85th percentile) based on the WHO reference values for BMI-for-age (WHO, 1995). The discrepancy between actual and perceived body size was determined and each subject was rated as either having a correct perception, an underestimation of their size (ie they were bigger than they thought) or an overestimation of their true size (ie they were thinner than they thought).

Since all the outcome variables were categorical, differences by gender and ethnicity were analyzed using the Chi-squared test. The level of statistical significance was set at P<0.05. The data were analyzed using SPSS 9.0 for Windows.

Results

A total of 1139 students were selected for study and 1090 (95.7%) consented and participated, of whom 578 (53%) were female. The age ranged from 14 to 17 y and the mean was 15.3 y. The age range was wide (14–17 y), a reflection of the education policy in Trinidad where entry to secondary school is related to readiness and not simply age. The main ethnic groups were South Asians (49%), African (25%) and persons of mixed ethnicity (23%).

The actual, perceived, discrepancy between actual and perceived and most attractive (of the opposite sex) body sizes are shown, by ethnicity, in Table 1. Overall, 15% of the sample was thin while 13% was overweight/obese. There was a significant ethnic difference in actual body size (P<0.001). There were more thin adolescents of South Asian descent than of African or mixed ethnicity. This ethnic difference was present among both the male and female subjects. The only gender difference in actual body size was among the South Asian group with more thin males (28%) than females (18%; P=0.002).

Table 1 - Actual, perceived, discrepancy between actual and perceived, and most attractive (of the opposite sex) body size, by ethnicity.

Full table (31K)

Overall, 15% of the adolescents perceived themselves as being thin, 65% normal, 18% overweight and 2% obese. There was no difference in perceived body size by ethnicity. When the discrepancy between actual and perceived body size was examined it was found that 68% correctly identified their body size while 13 and 19% underestimated and overestimated their actual size, respectively. The ethnic difference was only significant among the males with more South Asians (35%) over-estimating their size than the Africans (17%) or males of mixed ethnicity (22%; P=0.001). There were gender differences among the adolescents of South Asian (P<0.001) and mixed ethnicity (P=0.01). In each case, more males than females thought that they were bigger than they really were.

There was an ethnic difference in the most attractive body size but it was only significant among the males (P<0.001). Fewer of the South Asian males found the overweight female silhouette to be the most attractive (2%) compared with the African (14%) or mixed (13%) males. There was a significant gender difference in the most attractive body size (P<0.001), which was consistent among all the ethnic groups. More girls (28%) than boys (9%) thought that the overweight size was the most attractive.

Satisfaction with their current body size and plans for weight change are shown by ethnicity, in Table 2. There was an ethnic difference in satisfaction with body size, but it was only significant among the females (P=0.04). Fewer girls of mixed ethnicity (53%) were satisfied with their current size than the Africans (63%) or South Asians (65%). There was no gender difference in satisfaction for any of the ethnic groups. Overall, 68% of males and 61% of females were satisfied with their size. When satisfaction was examined by current body size, there were interesting ethnic differences. Seventy-one percent of the thin South Asians were satisfied with their size compared with only 46% of the thin Africans and 43% of thin persons of mixed ethnicity (P=0.04). Also, 48% of the overweight Africans were satisfied with their size compared with 27% of the overweight South Asians and 20% of the overweight persons of mixed ethnicity (P=0.03).

Table 2 - Satisfaction with body size and desire for weight change, by ethnicity.

Full table (11K)

In the response to the question on any plans for weight change, there was also an ethnic difference but it was only significant in females (P=0.02). Fewer African females (11%) wanted to lose weight compared with those of South Asian (25%) and mixed ethnicity (25%). Also, whereas 41% of all subjects said that they would like to remain the same size, more females (37%) than males (20%) said that they wanted to lose weight while more males (38%) than females (23%) wanted to gain weight (P<0.001). This gender difference was present in all ethnic groups.

The students were asked which of the male and female silhouettes were associated with health, wealth and happiness (Table 3). Eighty-six percent of the sample chose the normal female body size as the healthiest while 66 and 32% chose the normal and overweight male figures, respectively. There was more congruence in the silhouettes chosen as the least healthy with 57 and 41% of all students selecting the obese and thin female silhouettes, respectively, while 52 and 46% chose the obese and thin males silhouettes, respectively. Of great concern was the finding that 76% of the students chose the overweight or obese male silhouettes as the wealthiest with 67% choosing the overweight or obese female silhouette. Also, while the majority of the adolescents chose the normal silhouette as the happiest, 40% chose the overweight or obese male silhouette with 23% choosing similar female body sizes. The only ethnic difference was in the assessment of the least healthy female silhouette (P=0.004). More South Asians (63%) chose the obese than the Africans (53%) or persons of mixed ethnicity (51%). There were gender differences in the assessment of the least healthy male and female silhouettes (P=0.01 and P=0.02, respectively). In each case, the males were more likely than the females to describe the thin silhouette as unhealthy while the females were more likely than the males to describe the obese silhouette as unhealthy.

Table 3 - Male and female silhouettes associated with health, wealth and happiness.

Full table (16K)

Discussion

Our survey of adolescents attending secondary schools in Trinidad identified very interesting ethnic differences in their perceptions of body size. Most important was probably the finding that overweight Africans were more likely to be satisfied with their size than the other overweight adolescents while thin South Asians were more likely to be satisfied with being thin than the other thin persons. In addition, the African girls were also less inclined to want to lose weight than the other girls. This may have great significance for the occurence of obesity when they become adults. Indeed, obesity has been recognized as a problem in the Caribbean (Martorell et al, 2000), with high rates being reported among adults of African origin, especially the women (Foster et al, 1993). The attitudes of the South Asians adolescents towards thinness were similar to those of white female adolescents in the USA.

We believe that it is important to correctly identify one's own size as persons with the correct perception of their body size are less likely to take inappropriate action, such as trying to lose weight when there is no reason. Overall, 68% of the adolescents were able to do this. The greater likelihood for South Asian males to overestimate their body size than the other adolescents may be related to the high rate of thinness in that group. As many as 28% of them were thin, which was similar to the results of another survey in Trinidad (Batson et al, 2001). It is likely that this is not an indication of the high rate of undernutrition in this group but rather a reflection of the inadequacy of the WHO reference data for this ethnic group (de Onis et al, 2001). The appropriateness of using of international reference data to determine the nutritional status of South Asians has been debated for some time (Messer, 1989) and has led to suggestions for a different BMI cut-off point to indicate thinness or undernutrition from that used for other groups (Naidu & Rao, 1994). More relevant to the current situation in the Caribbean was the finding that 13% of the sample were overweight or obese. This is much lower than in developed countries such as the USA (Styne, 2001). However, given the high rate of adult obesity among adults in the Caribbean, this should not deter plans for intervention programs, which should begin from this age group, if not earlier.

Interestingly, more than a quarter of the female students thought that the overweight male silhouette was the most attractive while the majority of the males chose the normal size female. This was consistent with our previous finding that over 30% of the males wanted to be overweight (Simeon et al, 2001). However, it is difficult to interpret this observation as it is not clear whether the students had regarded the 'overweight' male as having excess muscle or fat. The use of silhouettes was not able to communicate this distinction in body composition in the present study. Consequently, these findings may not necessarily indicate that we should expect a preponderance of obese males in the future. Also, such a situation would be very different from the current situation as the prevalence of obesity is much higher in females than males in the Caribbean (Foster et al, 1993).

Assessing the body images that the adolescents associated with health, wealth and happiness also produced relevant and important findings. Many of them associated the overweight male with health and happiness, while the situation was not quite the same with the overweight female. It is also important to note that, although many persons associated obesity with poor health, the majority also associated it with wealth. It is clear that these students do not necessarily equate health with wealth. The association of obesity with wealth and happiness was previously found in a Barbados study comprising adults (Hoyos & Clarke, 1987). It was therefore instructive to note that young persons had similar beliefs. This will undoubtedly have serious implications for our drive to control obesity in the Caribbean.

The study only comprised adolescents attending school, it is therefore not possible to extrapolate the results to all adolescents in Trinidad. However, the majority of Trinidadians aged 14–17 y attend school. Silhouettes were used to assess perceptions of body size. While this visual aid was very useful and had been successfully used in other studies (Rand et al, 1997), there may have been difficulties in distinguishing body size from shape. Another limitation was that BMI was used to measure actual body size; however it does not measure the quantity and distribution of adiposity in an individual. This is particularly important in males where a high BMI can be due to either excess adiposity or muscle. In order to address this, it is recommended that skin-fold thicknesses be measured. However, these measurements were a great inconvenience to students in a pre-test and we felt that, if we had attempted to take them, the response rate would have been adversely affected. Consequently, we were not able to distinguish the obese from the overweight adolescents. In spite of these limitations we believe that, with the stratified random sampling technique that we used and the 96% response rate, our study was sound and provided valuable insights into the body image of adolescents in this multi-ethnic Caribbean population.

The ethnic differences that we found were likely to be related to culture; however further studies are needed to get a better understanding of this. Finally, although the prevalence of overweight/obesity was not high, the perception of the Trinidadian adolescent that obesity was associated with wealth and to a lesser extent happiness, and the satisfaction of the overweight African females with their size, were all causes for concern.

References

1. Abbott, J, Conway, S, Etherington, C, Fitzjohn, J, Gee, L, Morton, A, Musson, H & Webb, A ((2000)). Perceived body image and eating behavior in young adults with cystic fibrosis and their healthy peers. J. Behav. Med., 23, 501–517.
2. Batson, Y, Teelucksingh, S & Simeon, D ((2001)). Dietary patterns and nutritional status of secondary school children in Trinidad. W. Indian Med. J., 50, (Suppl 2)41
3. Borzekowski, DL, Robinson, T & Killen, J ((2000)). Does the camera add 10 pounds? Media use, perceived importance of appearance and weight concerns among teenage girls. J. Adolesc. Health, 26, 36–41.
4. de Onis, M, Dasgupta, P, Saha, S, Sengupta, D & Blossner, M ((2001)). The National Center for Health Statistics reference and the growth of Indian adolescent boys. Am. J. Clin. Nutr., 74, 248–253.
5. Fitzgibbon, ML, Blackman, LR & Avellone, ME ((2000)). The relationship between body image discrepancy and body mass index across ethnic groups. Obes. Res., 8, 582–589.
6. Foster, C, Rotimi, C, Fraser, H, Sundarum, C, Liao, Y, Gibson, E, Holder, Y, Hoyos, M & Mellanson-King, R ((1993)). Hypertension, diabetes, and obesity in Barbados: findings from a recent survey. Ethn. Dis., 3, 404–412.
7. Hoyos, MD & Clarke, H ((1987)). Concepts of obesity in family practice. W. Indian Med. J., 36, 95–98.
8. Huon, GF & Brown, L ((1986)). Attitude correlates of weight control among secondary school boys and girls. J. Adolesc. Health Care, 7, 178–182.
9. Kiefer, I, Leitner, B, Bauer, R & Rieder, A ((2000)). Body weight: the male and female perception. Soz. Praventivmed., 45, 274–278.
10. Killen, JD, Taylor, C, Hayward, C, Haydel, K, Wilson, D, Hammer, L, Kraemer, H, Blair-Greiner, A & Strachowski, D ((1996)). Weight concerns influence the development of eating disorders: a 4-year prospective study. J. Consult. Clin. Psychol., 64, 936–940.
11. Martorell, R, Khan, L, Hughes, M & Grummer-Strawn, L ((2000)). Obesity in women in developing countries. Eur. J. Clin. Nutr., 54, 247–252.
12. Messer, E ((1989)). Indian nutritionists and international nutrition standards: concepts and controversies. Soc. Sci. Med., 29, 1393–1399.
13. Miller, KJ, Gleaves, D, Hirsch, T, Green, B, Snow, A & Corbett, C ((2000)). Comparisons of body image dimensions by race/ethnicity and gender in a university population. Int. J. Eat. Disord., 27, 310–316.
14. Moore, DC ((1988)). Body image and eating behavior in adolescent girls. Am. J. Dis. Child., 142, 1114–1118.
15. Must, A, Jacques, PF, Dallal, GE, Bajema, CJ & Dietz, WH ((1992)). Long term morbidity and mortality of overweight adolescents: a follow-up of the Harvard Growth Study of 1922 to 1935. New. Engl. J. Med., 327, 1350–1355.
16. Naidu, AN & Rao, NP ((1994)). Body mass index: a measure of the nutritional status of Indian populations. Eur. J. Clin. Nutr., 48, (Suppl 3)S131–S140.
17. Rand, CS, Resnick, JL & Seldman, RS ((1997)). Assessment of socially acceptable body sizes by university students. Obes. Res., 5, 425–429.
18. Ricciardelli, LA & McCabe, MP ((2001)). Children's body image concerns and eating disturbance: a review of the literature. Clin. Psychol. Rev., 21, 325–344.
19. Robinson, TN, Killen, J, Litt, I, Hammer, L, Wilson, D, Haydel, K, Hayward, C & Taylor, C ((1996)). Ethnicity and body dissatisfaction: are Hispanic and Asian girls at increased risk for eating disorders?. J. Adolesc. Health, 19, 384–393.
20. Robinson, TN, Chang, J, Haydel, K & Killen, J ((2001)). Overweight concerns and body dissatisfaction among third grade children: the impacts of ethnicity and socioeconomic status. J. Pediatr., 138, 158–160.
21. Schreiber, GB, Robins, M, Striegel-Moore, R, Obarzanek, E, Morrison, J & Wright, D ((1996)). Weight modification efforts reported by black and white preadolescent girls: National Heart, Lung and Blood Institute Growth and Health Study. Pediatrics, 98, 63–70.
22. Simeon, DT, Abdool, P, Ali, A, Kungeesingh, K, Panchoo, K & Rattan, R ((2001)). Slim is not in' among adolescents in Trinidad?. J. Adolesc. Health, 28, 370–371.
23. Smith, DE & Gogswell, C ((1994)). A cross-cultural perspective on adolescent girls' body perception. Percept. Mot. Skills, 78, 744–746.
24. Stice, E, Hayward, C, Cameron, R, Killen, D & Taylor, C ((2000)). Body-image and eating disturbances predict onset of depression among female adolescents: a longitudinal study. J. Abnorm. Psychol., 109, 438–444.
25. Strauss, RS ((1999)). Self-reported weight status and dieting in a cross-sectional sample of young adolescents: National Health and Nutrition Examination Survey III. Arch. Pediatr. Adolesc. Med., 153, 741–747.
26. Striegel-Moore, RH ((2001)). Body image concerns among children. J. Pediatr., 138, 158–160.
27. Stunkard, AJ & Burt, V ((1967)). Obesity and the body image II: age at onset of disturbances in body image. Am. J. Psychiat., 123, 1443–1447.
28. Styne, D ((2001)). Childhood and adolescent obesity. Prevalence and significance. Pediatr. Clin. No. Am., 48, 823–854.
29. Taylor, CB, Sharpe, T, Shisslak, C, Bryson, S, Estes, L, Gray, N, McKnight, K, Crago, M, Kraemer, H & Killen, J ((1998)). Factors associated with weight concerns in adolescent girls. Int. J. Eat. Disord., 24, 31–42.
30. Tiggemann, M, Gardiner, M & Slater, A ((2000)). 'I would rather be size 10 than have straight A's': a focus group study of adolescent girls' wish to be thinner. J. Adolesc., 23, 645–659.
31. WHO ((1995)). Physical Status: the use and interpretation of anthropometry, WHO Technical Report Series no. 854.Geneva: WHO