Participants

The MyHeARTs study is a cohort study that recruited secondary school students from three States in the Central and Northern Regions of the Peninsular Malaysia, namely Selangor, Federal Territory of Kuala Lumpur, and Perak. Exactly 1361 students were recruited. The schools were stratified into urban and rural, based on criteria provided by the Department of Statistics, Malaysia.The dietary data used in this paper are from the first wave of this cohort, held between March and May, 2012. Details of the study were reported using published protocol [10], and were approved by ethics committees at the University of Malaya Medical Centre (MEC Ref. No: 896.34).

Our study included school children aged 13 years, in their first year at a public secondary school. As a requirement, the participants needed to read Malay, the national language of Malaysia. Other types of schools, like boarding, religious, and vernacular schools, were all excluded from out study, as the majority of students attending these schools encompass a single ethnic group. Participants and their parents or guardians received detailed and written information about the study as well as consent forms. Prior to participation, all participants and their parents or guardians were required to submit a completed consent form, thus indicating their willingness to participate. To categorize the education level of the parents, primary school was defined as education obtained in a school between the ages of 7 and 12 years—this was also called primary education; secondary school was considered as education obtained in a school between the ages of 13 and 17 years—this was also called secondary education; STPM, or a College or University level education, was defined as education obtained following the completion of secondary school—also called tertiary education. Based on family income, the socioeconomic status (SES) of the participants was categorised into low (<RM1500), middle (RM 1501–RM 5000) and high (>RM5000) SES [19].

Instruments and procedures

Dietary intake among adolescents by gender and its comparison with Malaysian RNI

Overall, males had a 10.1% higher mean energy intake compared to female adolescents. The majority of male energy intake was derived from carbohydrates, proteins, fats, and saturated fat; this finding was similar to other published studies, such as the Young Hearts Project [25], which reported that adolescent males aged 12 to 15-years consumed significantly more energy, compared to female adolescents (13.1 ± 3.58 MJ/d vs 8.4 ± 2.46 MJ/d); another study showed an increased intake of energy among adolescent males (mean, 12.9 ± 3.5 MJ/d) versus females [26]. Similar trends in total energy intake were found between boys and girls in other studies, such as the Young Hearts Project [25] and the Goteborg adolescent studies [26], yet the overall energy intake by the Malaysian adolescents participating in this study was lower than those two studies. Despite the 7-day diet history being used during the Young Hearts Project, the range of energy intake for adolescents aged 12 to 15 years was greater than the current study, which only involved adolescents aged 13-years. Older adolescents have different energy requirements compared to younger adolescents, which explains the range of energy intake. Findings on energy intake from the current study are comparable with another adolescent study conducted in Malaysia; [27] although the energy intake findings of the current study are comparable with this other study, it should be noted that we used different methods to investigate dietary intake—the Eating Behaviors Questionnaire and a three-day dietary record were used by the previous study. Because the current MyHeARTs study (2014) is an on-going longitudinal study, trends of energy and dietary intake by adolescents can be monitored.

Compared with the Malaysian RNI, the energy intake of the adolescents in current study reached an average of 65.5% (rural) and 66.3% (urban) for males, and 71.2% (rural) and 75.7% (urban) for females. Energy intake aside, male adolescents in the current study also showed higher carbohydrates, fats, cholesterol, saturated fat, MUFA and proteins intakes, compared with female adolescents (p-value <0.001). Comparing the actual macronutrient intake against the Malaysian recommended nutrient intake (RNI), we found that both male and female adolescents consumed adequate amounts of proteins, carbohydrates, fats, cholesterol and saturated fat. Even though female adolescents consumed adequate proteins, their iron intake did not fulfill RNI requirements. In line with the current study, another study in Malaysia reported similar results, whereby female adolescents only achieved 61% of their recommended iron intake [28]. Despite sufficient carbohydrates consumption, crude fiber intake among adolescents was found to be very low, and much lower than the Malaysian RNI (2.5 g/d vs 26-38g/d). This trend of inadequate macro and micronutrient intakes is similar to a cross-sectional study that observed 382 obese adolescents in the Kajang District, Selangor State; [27] in our study, both male and female adolescents showed inadequate energy and nutrient intake, including carbohydrates, fiber, vitamin D, calcium, and iron; because the focus of the study in Kajang District was only obese adolescents, the possibility of underreporting must be taken into consideration.

As for micronutrients, the overall sodium intake in our study (2289.5 ± 920.9 mg/d) exceeded the recommended intake, as necessitated by the Malaysian RNI (<2000 mg/d) [24,29]. In Malaysia, little is known of the current salt intake of adolescents. That said, a study by Moy, Gan, and Siti Zaleha (2006) and Zalilah et al. (2006) reported that ‘keropok’, which is a type of local snack that is high in salt and made from shrimp or fish mixed with flour, was a common snack consumed by adolescents [30,15]. Further investigation is needed to explore the types of foods taken by adolescents in a single meal; this is important, because it would help to educate and prevent adolescents from continuing with practices that may negatively impact their future health, practices that might aid the development of non-communicable diseases. Regardless of gender and geography, all adolescents consumed inadequate levels of calcium and vitamin D, when compared with the RNI. This finding was comparable to another study of adolescents in Malaysia [31], where similar trends were found in other countries, such as Korea [32], Brazil [33] and other European countries [34]. One study carried out in the State of Sabah, Eastern Malaysia, reported that calcium and iron intakes for adolescents aged 15 years was lowest compared to other nutrients, at 33.4% and 47.2% of the Malaysian RNI respectively [15]. In our study, calcium and vitamin D were consumed the least (< 50% of the RNI), for both males and females. These findings suggest a pro-active intervention is required if nutritional intake for adolescents to be optimized. Since 2006, Malaysia has adopted a “school milk programme” for school-aged children, yet surprisingly the uptake of milk and milk products among adolescents remains low, as was observed during this study. Since the number of foods fortified with vitamin D is minimal in Malaysia, our findings on Vitamin D intake are also not surprising.

Comparison of adolescents’ dietary intake by BMI and places of residence

Obese adolescents in the current study had a higher energy intake, 15.8% more than normal weight adolescents—and more than 30% of this energy intake by these obese adolescents came from fats intake (70.3 g/d). Contradicting the current findings, another study reported that the crude energy intake among the overweight class of adolescents was the lowest, compared to normal and underweight adolescent classes—noting that this other study was adjusted by body weight (p-value <0.001) [17]; it must be also noted that this other study used a 3-day self-reporting food record, and thus the divergent findings of our current study might be due to the different methods used to assess dietary intake. For the current study, adolescents from rural schools had a 5.5% increase of energy intake (1706 kcal/d) compared to those from urban schools (1612 kcal/d). Such findings are similar to a study undertaken in the Chapra District in the State of Bihar in rural India, where the average food intake was higher for adolescent males from rural areas [16]. Furthermore, we found that adolescents in rural settings consumed greater amounts of high-cholesterol foods, which was similarly observed in an Ecuadorian study [35]. For the rural contingent alone, our current study showed a significant difference in energy intake—carbohydrates and sugars—between the obese group and the normal weight group (results not shown): this indicates a disparity between the dietary intake of adolescents, in urban versus rural settings; this also justifies the need for an in-depth educational programme on nutrition, especially for adolescents in rural areas, to help them improve their food preferences and intake, and to help them minimize the risks of developing non-communicable diseases later in life.

The current study did not reveal a significant difference in sugar intake between adolescents from urban and rural jurisdictions, results that were different than the Ecuadorian [35] study, which found that sugar intake was higher among urban adolescents versus their rural counterparts, with a median added-sugar intake of 86.5 g/d (approximately 17 teaspoons/d). Another study from the United States, of children and adolescents, showed that sugar-sweetened beverage topped all other food and beverage options as the primary source of sugar intake [14,36]. Comparing sugar intake with BMI, the differences we observed in our study between obese and normal weight adolescents was higher, by 15.1 g/d, and was 30.9% higher when compared to the normal weight adolescent group (p-value < 0.001). The 2005 Malaysian RNI guidance limits adolescent sugar consumption to < 15% of the total carbohydrates allowance, while the 2003 WHO guidance limits sugar intake for adolescents to < 10% of the total carbohydrates allowance. Thus, obese adolescents in present study consumed relatively higher amounts of sugar either from food or beverages (approximately 10 teaspoons per day). Other studies have shown that high sugar intake in overweight or obese adolescents is strongly correlated with sugar consumed via sugar-sweetened beverages [36–39]. The Malaysian Adult Nutrition Surveys (MANS) revealed that 59% of Malaysians consumed sugar and sweetened condensed milk on a daily basis [40] which would affect their child’s sugar intake. Middle-income counties should follow the practices and methods of higher-income countries, as observed in the National Health and Nutrition Examination Survey 1999–2008 [41]; that study showed a reduction of sugar intake from 100.1 g/d to 76.7 g/d, albeit sugar intakes were still greater than the recommended RNI levels. This reduction may be explained by increased public interest regarding the rise and prevalence of obesity, by a low carbohydrate diet, and by a reduction in the amount of sugar-sweetened beverages sold in schools. Although sugar intake among obese adolescents in our current study was not as high as the sugar intake found in the 1999–2008 survey [41], Malaysian adolescent sugar intake still exceeded the recommended daily sugar intake for adolescents (25–40 g/d) as recommended by the Malaysian RNI. It would be interesting to see if these trends continued in local setups, especially after the implementation of a no sugar subsidy policy instigated in Malaysia in 2013; this information, regarding sugar intake, is yet to be identified, and could be studied via the dietary assessments undertaken during the second wave of MyHeARTs from 2014.

Strengths and limitations

This study has several strengths. Firstly, the 7-day diet history was conducted on a one-to-one basis, with each adolescent interviewed by a trained dietician. Next, the interviewers used several tools to assist their data collection: they used visual representations of household measurement tools—plate, cup, glass, bowl, ladle, spoon; they used a flip-chart that outlined common Malay, Chinese and Indian foods; they used a diet history collection tool [19,26,42]. A 7-day diet history was used because a 24-hours diet recall alone will not adequately capture the precise diet of each participant [43]; a 7-day diet history also gives a better view of emerging dietary patterns, especially variations in macronutrient intake [44]. Furthermore, when studying younger persons, it has been shown that better energy intake estimates can be obtained at the group level over than the weighed record method; the weighed record method is also more expensive than analysis at the group level; furthermore, the weighed record method is not suitable for population-based studies [44].

By using a 7-day diet history, the habitual intake of our adolescent cohort was captured [26,43,45]; most importantly, this 7-day diet history is better suited to overcoming the age-related bias that is a typical of normal food intake reporting [46]. Our study is the largest study of adolescent dietary intake in Malaysia. Dietary data obtained from our initial cohort will improve the success of future dietary planning, and will help with current interventions, undertaken by caretakers, schools and by adolescents themselves, as each works to improve the general health of our students.

The time taken to conduct each interview, approximately 30 minutes or more per adolescent, can be reduced going forward. In the future, and during the 2016 MyHeARTs study specifically, a validated food frequency questionnaire will be used to help minimize the time required to collect dietary information. Furthermore, participant recall bias is one of the main areas of error during a dietary assessment, and is a common problem of all dietary assessment research, across all ages. However, this was minimized during the current study, and can be minimized during any dietary assessment research, by using qualified and trained dietitians to conduct the interviews.