In this study, approximately 4.6% of all adolescents and 11.8% of overweight/obese adolescents met the criteria for the metabolic syndrome. After adjusting for other relevant study factors (age, gender, pubertal status, maternal education and socio-economic status), there was a graded negative association between metabolic syndrome and moderate and vigorous physical activity. The odds of having metabolic syndrome rose in the second to the forth quartile of time spent on moderate to vigorous physical activity compared with the most active quartile. Furthermore, the odds of having metabolic syndrome were significantly higher in adolescents from higher socioeconomic families.
To date no study in Vietnam has examined the metabolic syndrome in adolescents or its relationship to physical activity. Our study provides evidence for a dose-response relation between physical activity and the risk of metabolic syndrome, and makes an important contribution to understanding the metabolic syndrome in adolescents in Ho Chi Minh City, where overweight and obesity are rapidly becoming a public health problem.
Our results revealed a high prevalence of metabolic syndrome among adolescents from Ho Chi Minh City compared to other countries in Asia. The prevalence observed from the 2002 China National Nutrition and Health Survey of 2761 adolescents aged 15 to 19 years was 3.7%, and was slightly higher in urban adolescents (4.2%) . A study from Japan reported a lower rate of only 1.4% . When the IDF criteria were applied in other studies, an even lower prevalence of metabolic syndrome was found in Chinese (1.2%)  and South Korean (1.8%) adolescents aged 10 to 16 years . Our findings indicate that the prevalence of adolescent metabolic syndrome in the urban population of Ho Chi Minh City was similar to that reported in the United States (4.2%) .
Using the same criteria, the prevalence of metabolic syndrome found among overweight and obese Vietnamese adolescents (9.5% and 17.1%, respectively) was higher than that were reported for South Korean adolescents (1.5% and 14.7% for overweight and obesity in 2005) , and for mildly obese and obese Japanese males (1.3% and 15.6%) . However, the prevalence of metabolic syndrome among overweight and obese Vietnamese adolescents was lower than that of overweight and obese Chinese adolescents (23.4% and 35.2%)  and that of overweight US adolescents (29%) , and moderate to severely obese US adolescents (39% and 50% respectively) .
A number of difficulties were encountered when trying to compare the findings from our study with the results from other countries. The estimates of the prevalence of metabolic syndrome from different countries were based on different reference data, different cut off values, or different criteria making cross country comparisons difficult. A further difficulty was the limited number of surveys assessing metabolic syndrome in adolescents from Asia, especially in urban populations as in our study.
The most common individual component of the metabolic syndrome in Vietnamese adolescents was elevated BP, followed by hypertriglyceridemia. This contrasts with the Korean studies that found glucose intolerance was the dominant feature of the metabolic syndrome . Among Turkish and Chinese adolescents, either low HDL-c and/or high TG levels were the most common components, followed by high BP [5, 35]. Similarly, an Indian study on metabolic syndrome of adolescents showed that low high-density lipoprotein was the most common and abdominal obesity the least common component of the metabolic syndrome . This heterogeneity suggests that we should take into account the role of genetic and ethnic factors when assessing the different components of metabolic syndrome.
The overweight/obese children in our study had significantly higher odds of having the metabolic syndrome compared to the non-overweight/obese counterpart. This finding is consistent with the observation that in children, obesity is strongly associated with metabolic syndrome [10, 11]. After adjustment for relevant factors, there was an inverse association between physical activity and metabolic syndrome. This result is also consistent with observations from other studies evaluating the impact of physical activity on the metabolic syndrome in youth [13–18]. Although cause and effect remain unclear from these findings, it seems that many adolescents in Ho Chi Minh City might benefit from spending more time in moderate to vigorous physical activity, which could possibly minimize their risk of developing metabolic syndrome.
The beneficial effect of physical activity for metabolic syndrome could be explained by its influence on body composition. Skeletal muscle is the most insulin sensitive tissue in the body. Physical activity has been shown to improve skeletal muscle insulin sensitivity and reduce insulin resistance by improving glucose transport in muscle cells and improving peripheral microcirculation [36, 37]. Additionally, an increased substrate use with a decrease of carbohydrate oxidation , as well as heightened insulin sensitivity  may play a role in the protective adaptations triggered by physical activity on metabolic risk factors.
It has been emphasized that a sustained amount of moderate to vigorous physical activity on a regular basis can induce and maintain beneficial effects on metabolic syndrome in adolescents . Furthermore, we can also find evidence which shows that physical activity is associated with improvement in elements of the metabolic syndrome, such as lower fasting insulin and greater insulin sensitivity in children  and reduction in LDL-c and diastolic BP .
Many studies have revealed that the prevalence of the metabolic syndrome significantly increases with increasing socio-economic status in both developing [35, 42] and developed countries . Socioeconomic status can affect diet and, hence, BMI and serum lipid levels in adolescents. Furthermore, in Vietnam, children in the wealthiest families were least likely to be active because their parents usually provide them with a "modern" life, including up-to-date recreational facilities such as televisions, computers, and other technical household devices and helpers that reduce the level of activity needed for daily household chores. Besides, wealthier families are more likely to be able to purchase ample food including energy dense foods and drinks.
It was also mentioned that the prevalence of metabolic syndrome increased when age increased regardless of gender [44, 45]. In the present study, the significance of the association between age and the metabolic syndrome was borderline (p = 0.05). That might be explained by the small number of children with metabolic syndrome found in this study, which resulted in limited power to detect significant associations.
There are conflicting findings from previous studies about the association of pubertal status and metabolic syndrome in adolescents. While studies have found no relationship between pubertal status and the likelihood of metabolic syndrome in adolescents [2, 16], others have revealed a statistically significant association between pubertal status and increased odds of metabolic syndrome among adolescents . We did not find a significant association between pubertal status and the odds of metabolic syndrome.
Similarly, we also did not find an association between gender and metabolic syndrome. Although, there seemed to be some gender differences with regard to the different components of the metabolic syndrome, these were all non significant, which might have been due to the somewhat limited statistical power of the study.
The key strengths of this study included the representative study sample of adolescents from Ho Chi Minh City that improved external validity of the study and the use of an objective measure of physical activity with accelerometers. We also used an age-appropriate measurement tool to identify metabolic syndrome in these adolescents.
A major limitation was the lack of any time dimension in assessing factors associated with metabolic syndrome due to the cross-sectional design of the study. Because both the outcome and the risk factors were examined at one point in time, we do not know whether these factors preceded or followed the onset of metabolic syndrome. An example of the inability of the study design to determine between cause and effect is the relationship between obesity and physical inactivity. Another limitation of this study was that the association between sedentary behaviors, including the time spent watching television and playing video and computer games, and the metabolic syndrome was not examined. Furthermore, the sample size of the study might have been a limitation as well, evidenced by the large confidence intervals for example of the moderate/vigorous physical activity odds-ratios for metabolic syndrome. Despite these limitations, our findings contribute to understanding the association between physical activity and the metabolic syndrome. These findings suggest that many adolescents in Ho Chi Minh City might need to spend more time on moderate to vigorous physical activity to minimize their risk of developing the metabolic syndrome.