Short report

Social and behavioural determinants of body mass index among adolescent females in Croatia

For females, body mass index (BMI; weight/height²) is the strongest positive predictor of body dissatisfaction and anxiety about physique.¹ Social factors could influence the relationship between BMI and weight-related behaviours, and mechanisms depend on sociodemographic group differences.²

This cross-sectional study evaluated the influence of sociobehavioural factors on BMI in female adolescents in Trogir, Croatia. We hypothesized that BMI in young females is associated with health behaviours and life style.

During April 2001, 328 females (aged 15.78±1.19 years, median 15.83 years, range 13.31–18.10 years) who attended a comprehensive school in Trogir, Croatia returned a self-report questionnaire regarding life style and health behaviours. After data collection, participants were subjected to anthropometric measurements.

Primary outcome measures were: body weight (kg); body height (cm); satisfaction with actual body weight (yes or no); following a diet regimen (yes or no); routinely or occasionally skipping meals (breakfast, lunch and dinner; yes or no); daily milk consumption (dl); age at menarche (years); success in school, compared with average school marks in Croatia (excellent, very good, good, sufficient and insufficient); sport activity (yes or no); activity outside school (yes or no); and smoking (yes or no, and number of cigarettes smoked per day). The secondary outcome measure was BMI.

Anthropometric measurements (body height and weight) were obtained³ by a trained clinician using a digital anthropometric device (Secca 220, Hamburg, Germany). Biases of the device were ±0.01 kg and ±5 mm.

The results were expressed as arithmetic mean±standard deviation. Normality of distribution was tested with the Shapiro–Wilk's W-test. Differences between two groups were calculated using Student's t-test for independent data (two-tailed). Correlations between variables were tested using Pearson's or Spearman's correlation coefficients (one-tailed). A P value of less than 0.05 was considered to be statistically significant.

Overall, the mean BMI was 20.59±2.26 kg/m² (median 20.29 kg/m², range 14.64–30.08 kg/m²). Subjects who were satisfied with their body weight (n=126) had lower BMI values than subjects who were dissatisfied (19.37±1.25 kg/m² vs 21.37±2.42 kg/m², P<0.001).

There were significant differences in BMI values between dieting females (n=56) and non-dieting females (21.47±2.36 kg/m² vs 20.41±2.20 kg/m², P=0.002).

Table 1 shows the differences between subjects who occasionally or routinely skipped breakfast/dinner and subjects who ate breakfast/dinner. There was no difference in BMI between the subjects who occasionally or routinely skipped lunch (n=95) and those who ate lunch.

There was no difference in BMI between subjects who consumed at least 1 dl of milk each day (n=264, 80.5%) and those who did not drink milk. The daily quantity of milk consumption (mean 4.36±3.47 dl) was correlated (Pearson's correlation coefficient, one-tailed) with weight (r=−0.122, P=0.015) and BMI (r=−0.102, P=0.035).

Age of menarche (mean 12.90±1.10 years, median 13.00 years, range 11.00–16.50 years) was positively correlated with height (r=0.154, P=0.003) and negatively correlated with BMI (r=−0.136, P=0.008).

Correlations between outcome measures and success in school (Spearman's correlation coefficient with significance, one-tailed) were demonstrated for body weight (ρ=−0.121, P=0.015) and height (ρ=−0.110, P=0.025). When subjects were categorized as ‘excellent pupils’ (n=98) or ‘others’, significant differences in body weight (57.62±6.05 kg vs 59.60±7.94 kg, P=0.034) were seen between the groups.

Outcome measures in subjects who took part in sporting activities (n=195) did not differ significantly from inactive subjects. The same result was obtained in subjects who took part in activities outside school (n=215).

No significant difference in BMI was found between smokers (n=85) compared with non-smokers, and there was no correlation between the number of cigarettes smoked per day (mean 10.75±7.04, range 1–40) and the outcome measures.

This study demonstrated that BMI among this group of adolescent females could be influenced by many factors. Dissatisfaction with body mass was associated with higher BMI, and this observation demonstrated a relationship between BMI and body perception among this study group. Furthermore, dieters had higher BMIs than non-dieters. Hausenblas et al.⁴ found that BMI was positively related to body image measures in females.

Success in school was negatively associated with body weight; excellent pupils had lower body weight compared with other pupils. This observation is mainly the result of social influence and the socially imposed template of ‘successful and slim women’. The female adolescents in this study were obviously receptive for such social pressure. Around the world, a slim figure is the socially acceptable body shape for female adolescents and young women. In particular, Western women have a tendency to believe that they are fat, even if they are classified as thin or ‘normal’.⁵

Skipping meals had a significant influence on body weight. Skipping dinner and/or breakfast was associated with a higher BMI. This could be because individuals with a higher BMI were skipping meals in an attempt to reduce body weight. Alternatively, the higher BMI could be due to skipping meals, as it has been reported previously that children who do not eat dinner regularly eat sweets and fast foods more often than children who eat dinner regularly.⁶

Daily quantity of milk consumption was associated with lower weight and BMI, and this could reflect the healthier life style of the milk consumers. Furthermore, Zemel et al.⁷ demonstrated that increasing dietary calcium suppresses adipocyte intracellular Ca²⁺ and thereby modulates energy metabolism and attenuates obesity risk.

Our data revealed that late menarche was associated with tallness and lower BMI, as expected. Some previous reports have shown that girls who undergo menarche at a relatively young age tend to be more obese as adults.⁸

Sport and activities outside school did not influence BMI. Donnelly et al.⁹ found that moderate-intensity exercise sustained for 16 months was effective for weight management in young adults.

Smoking for weight control is prevalent across many race/ethnic groups and in both genders during adolescence. Winter et al.¹⁰ concluded that body weight perceptions and the use of weight control behaviours were significantly associated with predictors of smoking among adolescent females. On the contrary, among this study group, self-reported smoking was not correlated with BMI.

In summary, this study provided an interesting insight into the BMI correlates in a group of female adolescents in Croatia. The results showed that females with higher BMIs demonstrated body mass dissatisfaction, weaker school success, meal skipping, dieting behaviour, less milk consumption and earlier menarche. BMI was not affected by sport, activities outside school or cigarette smoking.

Further studies, including a national Croatian study, are needed to increase our understanding of these findings.