Hostname: page-component-586b7cd67f-l7hp2 Total loading time: 0 Render date: 2024-11-26T09:29:26.101Z Has data issue: false hasContentIssue false

Energy intakes, macronutrient intakes and the percentages of energy from macronutrients with adolescent BMI: results from a 5-year cohort study in Ho Chi Minh City, Vietnam

Published online by Cambridge University Press:  10 October 2022

Hong K. Tang*
Affiliation:
Department of Epidemiology, Faculty of Public Health, Pham Ngoc Thach University of Medicine, Ho Chi Minh City, Vietnam
Ngoc-Minh Nguyen
Affiliation:
Department of Epidemiology, Faculty of Public Health, Pham Ngoc Thach University of Medicine, Ho Chi Minh City, Vietnam
Michael J. Dibley
Affiliation:
Sydney School of Public Health, Sydney Medical School, The University of Sydney, NSW 2006, Australia
*
*Corresponding author: Hong K. Tang, email [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Background:

Adolescence is a period of life when dietary patterns and nutrient intakes may greatly influence adult fatness. This study assesses the tracking of energy and nutrient intakes of Ho Chi Minh City adolescents over 5 years. It explores the possible relationships between energy and the percentage of energy from macronutrients with BMI.

Methods:

Height, weight, time spent on physical activity, screen time and dietary intakes were collected annually between 2004 and 2009 among 752 junior high school students with a mean age of 11·87 years at baseline. The tracking was investigated using correlation coefficients and weighted kappa statistics (k) for repeated measurements. Mixed effect models were used to investigate the association between energy intakes and percentage energy from macronutrients with BMI.

Results:

There were increases in the mean BMI annually, but greater in boys than in girls. Correlation coefficients (0·2 < r < 0·4) between participants’ intakes at baseline and 5-year follow-up suggest moderate tracking. Extended kappa values were lowest for energy from carbohydrate (CHO) in both girls and boys (k = 0·18 & 0·24, respectively), and highest for protein in girls (k = 0·47) and fat in boys (k = 0·48). The multilevel models showed the following variables significantly correlated with BMI: CHO, fat, percentage of energy from CHO, fat, time spent for moderate to vigorous physical activity, screen time, age and sex.

Conclusions:

The poor to fair tracking observed in this cohort suggests that individual dietary patterns exhibited in the first year are unlikely to predict energy and nutrient intakes in the fifth year.

Type
Research Article
Copyright
© The Author(s), 2022. Published by Cambridge University Press on behalf of The Nutrition Society

The prevalence of obesity in children and adolescents in Ho Chi Minh City, Vietnam, has been increasing significantly since the early 2000s. From 2002 to 2004, the prevalence of overweight among junior high school students aged 11 to 14 years doubled from 5·0 % to 11·7 %, while obesity prevalence increased threefold from 0·8 % to 2·0 %(Reference Hong, Dibley and Sibbritt1). Despite implementing health promotion programmes to prevent obesity(Reference Le, Do and Nguyen2,3) , the prevalence of overweight and obesity in the same population of adolescents still went up to 17·8 % and 3·2 % in 2013(Reference Nguyen, Hong and Hoang4). Adolescent obesity increases the risk of adult obesity, and evidence suggests that most overweight adolescents become overweight adults(Reference Dietz and Gortmaker5Reference Llewellyn, Simmonds and Owen7). Thus, there should be an investigation to understand and prevent overweight and obesity in adolescents and limit increases in overweight and obesity in adults in the years ahead.

Adolescence is a period of life when dietary patterns, eating behaviours and nutrient intakes may greatly influence adiposity and the risk of chronic diseases decades later(Reference Schneider, Dumith and Orlandi8Reference McNaughton, Ball and Mishra10). Specifically, in a recent systematic review(Reference Schneider, Dumith and Orlandi8), Bruna et al. compiled twenty-one studies that investigated the longitudinal relationship between diet and body fat in adolescence and early adulthood. Although they found insufficient evidence on the link between diet and body fat in this population, their findings suggest that the consumption of unhealthy food or food groups (higher energy density and lower amount of nutrients) appears to be associated with a higher quantity of body fat in adolescence and early adulthood. Also, they have highlighted the need for further research on the longitudinal relationship between food consumption and body fat in this young population. Also, the macronutrient composition of the diet (protein and fat) may play an important role in childhood obesity(Reference del Mar Bibiloni, Tur and Morandi11,Reference McGloin, Livingstone and Greene12) . We need to track the nutrient intakes from adolescence into young adulthood to assess the maintenance of dietary behaviours and the relative position in the rank of these behaviours over time. These routines formed in early life may influence the progression of chronic diseases later(Reference McNaughton, Ball and Mishra10,Reference Waterland and Jirtle13) . Interestingly, several longitudinal studies have investigated the correlation of nutrient intake from adolescence itoyoung adulthood, and the results have been inconsistent(Reference Schneider, Dumith and Orlandi8).

Obesity results from an imbalance in the energy balance equation(Reference Weihrauch-Blüher and Wiegand14,Reference Kumar and Kelly15) , hence, the rationale for investigating the relationship between energy intake and percentage energy intake from macronutrients with adolescent adiposity. Several studies assessing these relationships showed conflicting results, yet most were cross-sectional studies(Reference Maillard, Charles and Lafay16Reference Lotfi, Mohammadi and Mirzaei18). Thus, this longitudinal study and analysis can be beneficial. This paper aims to assess the tracking of energy and macronutrient (carbohydrate (CHO), lipid and protein) intakes of adolescents in a 5-year cohort study in Ho Chi Minh City. Our specific aims are to assess the longitudinal changes and relationships between energy and percentage of energy from macronutrients with BMI in adolescents.

Methodology

Subjects

This cohort study began from a multi-stage cluster cross-sectional survey in 2004 with thirty-one selected schools, including seventeen from wealthy urban districts and fourteen from less wealthy urban districts (n = 3319) in Ho Chi Minh City, Vietnam. We applied systematic random sampling to select eighteen schools from these thirty-one schools as the sub-sample for the cohort study, of which eleven were from wealthy districts and seven were from less wealthy districts. Then, one class from either grade 6 or 7 was selected in each school, resulting in 784 students being invited to join the cohort study. This study is a secondary analysis, and more details of the cohort design and results are published elsewhere(Reference Trang, Hong and Dibley19,Reference Trang, Hong and Van der Ploeg20) .

The Research Ethics Committee, Pham Ngoc Thach University of Medicine, HCMC, approved the cohort study. Participation in the study required the consent of both adolescents and their parents.

Data collection

We measured the main outcome variables (anthropometric measurements) and the exposure factors annually at each assessment round occurred at the beginning of the school year. Standing height was measured with a suspended Microtoise tape using standard methodology (precision of 0·1 cm). Body weight was measured using a Tanita BF 571 electronic scale, Tanita Corporation (precision to 0·1 kg). Information on physical activity and screen time was measured using the validated Vietnamese-Adolescent Physical Activity Recall Questionnaire (V-APARQ).(Reference Trang, van der Ploeg, Hardy and Dibley21). A validated Youth Food Frequency Questionnaire (YFFQ).(Reference Hong, Dibley and Sibbritt22) was used to assess dietary intake and diet behaviours. The YFFQ included questions regarding the usual frequency of intake of 160 specific food items over the last 6 months. We used EIYOKUN v.1.(Reference Patterson, Warnberg and Kearney23,Reference Hanh24) , a nutrient database developed from Vietnamese food consumption tables, to calculate foods’ energy, protein, CHO and fat content. Total daily nutrient intakes were then calculated by summing nutrient values for the individual foods from the frequency of consumption, the mean amount consumed and the nutrients per gram. We excluded implausible energy intakes < 500 kcal/d or > 5000 kcal/d.(Reference Rockett, Breitenbach and Frazier25). We calculated BMI as the weight (kg) divided by height squared (m2).

We assessed pubertal development by a self-administered confidential questionnaire using diagrams illustrating the five Tanner stages of pubertal development(Reference Tanner26). Also, we asked female students the date of the first menstruation and the male students the date of attaining the adult voice. Based on the WHO definition, the pubertal stages were grouped into pre-pubescent, pubescent and post-pubescent(27).

Data analysis

We report the results as mean and standard deviation for normally distributed continuous variables. We have presented the median and interquartile ranges, because most of the nutrient distributions were right-skewed towards higher values. To test differences in the mean of energy intakes and percentage energy from macronutrients, including protein, CHO and fat, between the first and the last year, we used paired t tests, or the Wilcoxon signed rank tests when data were not normally distributed. We examined the tracking between dietary intake measured at baseline and the fifth follow-up year by computing correlation coefficients using the mixed effects model suggested by Guogen et al. (Reference Shan, Zhang and Jiang28). The cut-off value of 0·2 was used to indicate tracking(Reference Li and Wang29). We also divided energy intakes and macronutrient intakes into tertiles: lowest tertile (L1), middle tertile (M1) and highest tertile (H1) and calculated the extended kappa statistics (k) for these categories based on the approach to deal with repeated measurements(Reference Ma, Tang and Feng30,Reference Pan, Rose and Haber31) . As suggested by Twisk JW et al(Reference Twisk, Kemper and Mellenbergh32) if k > 0·75, the variable tracks well, if k < 0·40, the variable tracks poorly, and if 0·40 ≤ k ≤ 0·7 then there is moderate tracking for the variable.

We examined the relationship between changes (from year 1 to year 5) in total energy intakes, protein, CHO, fat intakes, percentage energy from these three macronutrients, time spent for moderate to vigorous physical activity, total screen time, sex, age and maturation with changes of BMI. Multilevel mixed effects models were used to consider the cluster effect of schools, class and within-subject correlation from repeated measurements. We conducted all analyses using Stata version 15.0 (Stata Corporation, 2018). As recommended for longitudinal data analysis(Reference Berkey and Colditz33), we used BMI instead of BMI z-score.

Results

At baseline, 759 out of 784 students (selected from the cross-sectional study) consented to participate in the cohort study. We excluded seven observations due to mean energy intakes being < 500 kcal per d or > 5000 kcal per d, yielding 752 students with dietary intake data at the baseline. After the 5-year follow-up, we excluded seventy-six students due to the over-range of energy intake limits. Finally, 533 students remained for analysis. A total of 143 participants were not accounted for in the analysis as they were lost to follow-up. This cohort study had a high retention rate (70·9 %). We describe the characteristics of participants at the baseline and the last year of follow-up in Table 1. There were significant differences in BMI, protein, CHO, fat, and energy intakes between boys and girls (P < 0·001 for BMI and P < 0·05 for protein, CHO, fat and energy intakes). Between year 1 and year 5, the absolute increases in BMI were significant in both sexes (20·1 kg/m2 to 23·5 kg/m2 for males and 18·1 kg/m2 to 20·6 kg/m2 for females). Boys had higher BMI values than girls (P < 0·001). Every year, there were increases in the mean BMI; however, the gains were greater in boys than in girls.

Table 1. Participants’ physical characteristics as well as energy and macronutrient intakes by sex in 2004 and 2009

* CHO, Carbohydrate.

** MVPA, Moderate to Vigorous Physical Activity.

P-values of paired t tests.

P-values of Wilcoxon matched signed rank tests.

Table 1 also shows that boys had higher energy and nutrient intake. The median macronutrient distribution values were protein 14·7 %, CHO 50 % and fat 21·4 % of energy intake in boys in the first year, while in girls, these values were 15·0 %, 41·9 % and 22·8 %, respectively. For boys in the last year, the median macronutrient distribution values were protein 15·3 %, CHO 55·9 % and fat 23·9 %, while in girls, they were 16·4 %, 46·5 % and 24·6 %, respectively. Over the 5 years, boys and girls’ energy and macronutrient intakes increased significantly (P < 0·001) with increasing age. There were significant differences in macronutrient intakes and percentage of energy from macronutrients between sexes at baseline and year 5, except for fat intake at baseline and percentage of energy from fat in the last year, with boys consuming more than girls.

Table 2 presents the correlation coefficients and weighted kappa values for energy, macronutrient intakes and percentage of energy from macronutrients. Generally, the correlation between energy intakes and macronutrient intakes at the individual level was poor to fair but lower for CHO intakes. The weighted kappa values are fair for energy from CHO. In most cases, the weighted kappa values are higher in boys than in girls, especially with energy from CHO (Table 2).

Table 2. Tracking of energy and macronutrient intakes between baseline and the fifth follow-up year as estimated by correlation coefficients using a mixed effects model and weighted kappa statistics (k)

* CHO, Carbohydrate.

Weighted kappa statistics (k) on tertiles of the distribution.

In Table 3 and Table 4, we separately examine the association between BMI and each macronutrient and the association between BMI and the percentage of energy from each macronutrient. While protein intake was significantly associated with BMI in univariate analysis but changed to insignificant in the multivariate model, percentage energy from protein was not significantly associated with BMI in univariate analysis, thus not included in the multivariate model. We recognised that the direction of the univariate association of BMI with CHO and the percentage of energy from CHO flipped in the multivariate model (see Table 3 and Table 4). However, the negative association between CHO and the percentage of energy from CHO with BMI in the multivariate model is small (coefficient –0·0017 for CHO or –0·8439 for the percentage of energy from CHO).

Table 3. Mixed effects models between total energy intake, macronutrient intakes and time spent in MVPA, sedentary time, sex and age with adolescent BMI over 5 years

* CHO, Carbohydrate.

** MVPA, Moderate to Vigorous Physical Activity.

Multivariate data analysis using a mixed effects model includes the following variables: total energy intakes, sex, age, time spent for MVPA, total screen time and each macronutrient (protein or fat or CHO itself) with BMI.

Table 4. Mixed effects models between total energy intake, percentage energy from protein, fat and CHO and time spent in MVPA, sedentary time, sex and age with adolescent BMI over 5 years

* CHO, Carbohydrate.

** MVPA, Moderate to Vigorous Physical Activity.

Multivariate data analysis using a mixed effects model includes the following variables: total energy intakes, sex, age, time spent for MVPA, total screen time and percentage energy from CHO/or percentage energy from fat with BMI.

We checked for the sex differences in the multivariate models between BMI and CHO or percentage of energy from CHO and found that there was a difference between boys and girls: the association between BMI and CHO or percentage of energy from CHO was still significant for boys (the coefficient = −0·0030, with P-value = 0·013 and coefficient = −1·7544, with P-value = 0·028, respectively), but not significant for girls (the coefficient = −0·0008, with P-value = 0·245 and coefficient = −0·3838, with P-value = 0·348, respectively) (data were not shown).

We also examined the association with BMI for those who were not overweight or obese at baseline and got the same results as for the analysis of the total sample.

Discussion

Among 533 adolescents from 11 to 15 years old from across Ho Chi Minh City in our sample, at the group level, we found that energy and macronutrient intakes significantly changed throughout the study from 2004 to 2009. The correlation between energy and macronutrient intakes was poor to fair at the individual level, indicating a substantial drift of subjects between the low, medium and high intake categories with increasing age. As stated, the tracking coefficient highly depends on the length of time interval(Reference Twisk34); the 5-year period we looked at is long enough to observe changes in eating habits and stability of food preferences. Thus, we expected a low correlation for dietary intake(Reference Robson, Gallagher and Livingstone35) and much lower than for biological properties such as blood markers and anthropometric indicators(Reference Twisk, Kemper and van Mechelen36). Furthermore, the correlation for CHO and protein was the lowest but higher for fat intake, which at least two reasons might explain. First, the teenage students were increasingly influenced by their friends’ eating habits, which usually involved snacks and other fast foods (which also had a higher percentage of fat). They would switch to consuming more of these fast foods at school than homemade foods prepared by their parents (which have a more balanced nutrient content). Second, Vietnamese adolescents consume more energy from fat from ‘out-of-home’ foods(Reference Lachat, Khanh and Khan37), which keeps the percentage of fat intake and percentage of energy from fat higher than other macronutrient intakes. These reasons explain why fat intake increased. This increase over the 5 years was gradual, so the fat consumption data moved from one lower tertile to an adjacent higher one or stayed in the same high tertile, leading to a low correlation of nutrient intakes.

We expected that high intakes of fat and a high percentage of energy from fat would be associated with a high BMI. Findings from the present study do support this hypothesis. However, many studies worldwide have not shown a consistent relationship between energy and macronutrient intakes or the percentage of energy from macronutrients with BMI(Reference del Mar Bibiloni, Tur and Morandi11,Reference Lin, Mouratidou and Vereecken17,Reference Lotfi, Mohammadi and Mirzaei18,Reference Elliot, Truby and Lee38,Reference Dor, Stark and Dichtiar39) . While several studies have suggested that the macronutrient composition of the diet (protein, CHO and fat) plays an important contributing role in childhood obesity(Reference del Mar Bibiloni, Tur and Morandi11,Reference McGloin, Livingstone and Greene12,Reference Dor, Stark and Dichtiar39) , other studies have different findings(Reference Elliot, Truby and Lee38,Reference Hebestreit, Barba and De Henauw40) , or even an inverse relationship(Reference Lin, Mouratidou and Vereecken17,Reference Lotfi, Mohammadi and Mirzaei18) . In this study, we found a significant association between protein intake and BMI in the univariate analysis, but it became not significant in the multivariate model. A possible explanation for the lack of an association between protein intake or percentage energy from protein and BMI might be the misreporting of foods(Reference Huang, Howarth and Lin41,Reference Okubo and Sasaki42) . Our study excluded ‘implausible’ energy intakes from the analysis; however, this simple exclusion could not remove the effect of misreporting as this can happen in children within the included energy range but is more likely in the subgroup of children with higher BMI. When collecting data, we took two steps to minimise this bias; first, we spent a few minutes building trust and encouraging them to report as much food and drinks they possibly used; second, data collectors always checked if there were missing areas from the answered FFQ. However, to prevent or minimise this bias, we believe that better, or even more direct, tools intertwined in FFQ should be developed. In this study, we collected data among junior high school students exposed to unhealthy eating habits and sedentary behaviours. These students also started to have closer ties with their peers during the transition to adulthood. For them, the psychosocial importance of friendship and peer group relations increases gradually, thus intensifying the potential for peer influence(Reference Millstein, Petersen, Nightingale, Millstein, Petersen and Nightingale43). Therefore, adolescents’ food preferences/eating habits in Ho Chi Minh City may be less stable over time.

We noted that the direction of the univariate association of BMI with CHO and the percentage of energy from CHO has flipped in the multivariate model. A clear explanation for those negative associations is unsure. For further analysis, we also found sex differences in the multivariate models between BMI and CHO or the percentage of energy from CHO. Future research should explore interactions between the level of physical activity and sex and the dietary intake of CHO.

Our study showed a significant association between BMI with time spent on physical activity and screen time. Physical activity was negatively associated with BMI, while screen time was positively associated. The decreased physical activity and increased screen time among adolescents in this study are also consistent with findings reported in the USA, Europe and several other mid-income countries(Reference Farooq, Martin and Janssen44Reference Biddle, García Bengoechea and Wiesner46). We have published those results in several other papers(Reference Trang, Hong and Dibley19,Reference Trang, Hong and Van der Ploeg20) ; thus, they were not this paper’s focus.

The further analysis of the association of BMI of non-overweight or non-obese children at the baseline showed a significant relationship, as seen in the total sample. This analysis established that the baseline status (i.e. overweight or obese) did not influence or alter the current diet/eating behaviours in this study which is different from other reports(Reference Skinner, Steiner and Perrin47).

Our study has several strengths. Firstly, the longitudinal design allowed us to study changes over time in energy and macronutrient intakes and BMI while accounting for growth and maturation. BMI typically goes up from year 1 to year 5 among students in this cohort, and we took these changes into account. Secondly, we conducted the study among junior high school students who were followed closely for at least 3–4 years until they moved to different senior schools; hence, we had little loss to follow-up in the data. The retention rate was about 70·9 % and was not differential concerning baseline characteristics. Thirdly, we used various methods to assess the correlation patterns, including correlation coefficients, percentage agreement and kappa values, which allowed for comparisons and testing of the robustness of the related findings.

The main limitation of this study is that we gathered data from 2004 to 2009. Thus the information may be less relevant today due to the rapid changes in Vietnam over the past decade. Nonetheless, a recent study(Reference Nguyen, Hong and Hoang4) in 2013 showed similar patterns of overweight and obesity among adolescents in Ho Chi Minh, suggesting that our findings remain relevant. Similar reports from countries with the same economic situation as Vietnam found moderate or even poor to fair tracking in some dietary pattern scores(Reference Movassagh, Baxter-Jones and Kontulainen48,Reference Schneider, Dumith and Araújo de França49) . The second limitation is that data collected depended on memory and perception of the usual diet, and social desirability might influence responses. Hence, the degree of memory and motivation required to complete the FFQ may change substantially year by year, thereby contributing to the poor correlations observed in this study. Furthermore, we did not collect information on psychosocial factors likely to influence the changes in adolescents’ diets and BMI over time.

Despite these limitations, our findings indicate poor to moderate correlation patterns in the dietary intakes of urban adolescents over a 5-year follow-up. The poor to fair correlation observed in this cohort suggests that individual dietary patterns exhibited in the first year are unlikely to predict energy and nutrient intakes in the last year. This study also presents vital information on dietary intake and adiposity of adolescents in Ho Chi Minh City, which may prove useful in refining policies to prevent excess weight gain.

Acknowledgement

The Nestlé Foundation, Switzerland, funded a grant for the original cohort study. Ngoc-Minh Nguyen prepared with the help of a PhD scholarship from the Sydney Medical Foundation at The University of Sydney.

H. K. T. participated in the design, carried out the study, performed the statistical analysis and drafted the manuscript. N-M. N. contributed to the statistical analyses and drafted the manuscript. M. J. D. participated in the design of the research and the analytical strategy and contributed to drafting the manuscript. All authors read and approved the final manuscript.

There is no conflict of interest of any kind for any of the authors.

References

Hong, TK, Dibley, MJ, Sibbritt, D, et al. (2007) Overweight and obesity are rapidly emerging among adolescents in Ho Chi Minh City, Vietnam, 2002–2004. Int J Pediatr Obes 2, 194201.CrossRefGoogle Scholar
Le, TKQ, Do, TND, Nguyen, VT, et al. (2010) Effectiveness of nutrition intervention and physical excercise on overeweight/obesity control in primary school children in district 10 of Ho Chi Minh city in 2008–2009. J Food Nutr Sci 6, 13.Google Scholar
Centre HCMCN (2014) Ho Chi Minh City Implementing Effectively the School Nutrition Program. http://tphcm.chinhphu.vn/tphcm-trien-khai-hieu-qua-chuong-trinh-dinh-duong-hoc-duong (accessed February 2021).Google Scholar
Nguyen, PVN, Hong, TK, Hoang, T, et al. (2013) High prevalence of overweight among adolescents in Ho Chi Minh City, Vietnam. BMC Public Health 13, 141141.CrossRefGoogle ScholarPubMed
Dietz, WH & Gortmaker, SL (2001) Preventing obesity in children and adolescents. Ann Rev Public Health 22, 337353.CrossRefGoogle ScholarPubMed
Biro, FM & Wien, M (2010) Childhood obesity and adult morbidities. Am J Clin Nutr 91, 1499S1505S.CrossRefGoogle ScholarPubMed
Llewellyn, A, Simmonds, M, Owen, CG, et al. (2016) Childhood obesity as a predictor of morbidity in adulthood: a systematic review and meta-analysis. Obes Rev 17, 5667.CrossRefGoogle ScholarPubMed
Schneider, BC, Dumith, SC, Orlandi, SP, et al. (2017) Diet and body fat in adolescence and early adulthood: a systematic review of longitudinal studies. Ciência Saúde Coletiva 22, 15391552.CrossRefGoogle ScholarPubMed
Mikkilä, V, Räsänen, L, Raitakari, OT, et al. (2005) Consistent dietary patterns identified from childhood to adulthood: the Cardiovascular Risk in Young Finns Study. Br J Nutr 93, 923931.CrossRefGoogle ScholarPubMed
McNaughton, SA, Ball, K, Mishra, GD, et al. (2008) Dietary patterns of adolescents and risk of obesity and hypertension. J Nutr 138, 364370.CrossRefGoogle ScholarPubMed
del Mar Bibiloni, M, Tur, JA, Morandi, A, et al. (2015) Protein intake as a risk factor of overweight/obesity in 8– to 12–year-old children. Medicine 94, e2408.CrossRefGoogle ScholarPubMed
McGloin, AF, Livingstone, MB, Greene, LC, et al. (2002) Energy and fat intake in obese and lean children at varying risk of obesity. Int J Obes Relat Metab Disord 26, 200207.CrossRefGoogle ScholarPubMed
Waterland, RA & Jirtle, RL (2004) Early nutrition, epigenetic changes at transposons and imprinted genes, and enhanced susceptibility to adult chronic diseases. Nutrition 20, 6368.CrossRefGoogle ScholarPubMed
Weihrauch-Blüher, S & Wiegand, S (2018) Risk factors and implications of childhood obesity. Curr Obes Rep 7, 254259.CrossRefGoogle ScholarPubMed
Kumar, S & Kelly, AS (2017) Review of childhood obesity: from epidemiology, etiology, and comorbidities to clinical assessment and treatment. Mayo Clinic Proc 92, 251265.CrossRefGoogle ScholarPubMed
Maillard, G, Charles, MA, Lafay, L, et al. (2000) Macronutrient energy intake and adiposity in non obese prepubertal children aged 5–11 years (the Fleurbaix Laventie Ville Sante Study). Int J Obes Relat Metab Disord 24, 16081617.CrossRefGoogle Scholar
Lin, Y, Mouratidou, T, Vereecken, C, et al. (2015) Dietary animal and plant protein intakes and their associations with obesity and cardio-metabolic indicators in European adolescents: the HELENA cross-sectional study. Nutr J 14, 10.CrossRefGoogle ScholarPubMed
Lotfi, K, Mohammadi, S, Mirzaei, S, et al. (2022) Dietary total, plant and animal protein intake in relation to metabolic health status in overweight and obese adolescents. Sci Rep 12, 10055.CrossRefGoogle ScholarPubMed
Trang, NH, Hong, TK & Dibley, MJ (2012) Cohort profile: Ho Chi Minh City Youth Cohort--changes in diet, physical activity, sedentary behaviour and relationship with overweight/obesity in adolescents. BMJ Open 2, e000362.CrossRefGoogle Scholar
Trang, NH, Hong, TK, Van der Ploeg, HP, et al. (2012) Longitudinal physical activity changes in adolescents: Ho Chi Minh City Youth Cohort. Med Sci Sport Exerc 44, 14811489.CrossRefGoogle ScholarPubMed
, TK, Trang, NHHD, van der Ploeg, HP, Hardy, LL, Dibley, MJJ (2012) Validity and reliability of a physical activity questionnaire for Vietnamese adolescents. Int J Behav Nutr Phys Act 9, 93.CrossRefGoogle Scholar
Hong, TK, Dibley, MJ & Sibbritt, D (2010) Validity and reliability of an FFQ for use with adolescents in Ho Chi Minh City, Vietnam. Public Health Nutr 13, 368375.CrossRefGoogle ScholarPubMed
Patterson, E, Warnberg, J, Kearney, J, et al. (2009) The tracking of dietary intakes of children and adolescents in Sweden over six years: the European Youth Heart Study. Int J Behav Nutr Phys Act 6, 91.CrossRefGoogle ScholarPubMed
Hanh, TTM (2004) Eiyokun: Software for Calculating Vietnamese Dietary Intake. Ho Chi Minh City: Medical Publishing House of Vietnam.Google Scholar
Rockett, HR, Breitenbach, M, Frazier, AL, et al. (1997) Validation of a youth/adolescent food frequency questionnaire. Prev Med 26, 808816.CrossRefGoogle ScholarPubMed
Tanner, JM (1962) Growth at Adolescence, with a General Consideration of the Effects of Hereditary and Environmental Factors Upon Growth and Maturation from Birth to Maturity. 2nd ed. Oxford: Blackwell Scientific Publications.Google Scholar
WHO Expert Committee (1995) Physical Status: The Use and Interpretation of Anthropometry: Report of a WHO Expert Committee (WHO Technical Report Series no. 854). Geneva: World Health Organization.Google Scholar
Shan, G, Zhang, H & Jiang, T (2020) Correlation coefficients for a study with repeated measures. Comput Math Meth Med 2020, 7398324.CrossRefGoogle ScholarPubMed
Li, J & Wang, Y (2008) Tracking of dietary intake patterns is associated with baseline characteristics of urban low-income African-American adolescents. J Nutr 138, 94100.CrossRefGoogle ScholarPubMed
Ma, Y, Tang, W, Feng, C, et al. (2008) Inference for kappas for longitudinal study data: applications to sexual health research. Biom 64, 781789.CrossRefGoogle ScholarPubMed
Pan, Y, Rose, CE, Haber, M, et al. (2013) Assessing agreement of repeated binary measurements with an application to the CDC’s anthrax vaccine clinical trial. Int J Biostat 9, 1932.CrossRefGoogle Scholar
Twisk, JWR, Kemper, HCG & Mellenbergh, GJ (1994) Mathematical and analytical aspects of tracking. Epidemiologic Rev 16, 165183.CrossRefGoogle ScholarPubMed
Berkey, CS & Colditz, GA (2007) Adiposity in adolescents: change in actual BMI works better than change in BMI z score for longitudinal studies. Ann Epidemiol 17, 4450.CrossRefGoogle ScholarPubMed
Twisk, JW (2003) The problem of evaluating the magnitude of tracking coefficients. Eur J Epidemiol 18, 10251026.CrossRefGoogle ScholarPubMed
Robson, PJ, Gallagher, AM, Livingstone, MB, et al. (2000) Tracking of nutrient intakes in adolescence: the experiences of the Young Hearts Project, Northern Ireland. Br J Nutr 84, 541548.CrossRefGoogle Scholar
Twisk, JW, Kemper, HC, van Mechelen, W, et al. (1997) Tracking of risk factors for coronary heart disease over a 14-year period: a comparison between lifestyle and biologic risk factors with data from the Amsterdam Growth and Health Study. Am J Epidemiol 145, 888898.CrossRefGoogle Scholar
Lachat, C, Khanh, LNB, Khan, NC, et al. (2009) Eating out of home in Vietnamese adolescents: socioeconomic factors and dietary associations. Am J Clin Nutr 90, 16481655.CrossRefGoogle ScholarPubMed
Elliot, SA, Truby, H, Lee, A, et al. (2011) Associations of body mass index and waist circumference with: energy intake and percentage energy from macronutrients, in a cohort of australian children. Nutr J 10, 5864.CrossRefGoogle Scholar
Dor, C, Stark, AH, Dichtiar, R, et al. (2022) Non-dairy animal protein consumption is positively associated with overweight and obesity in Israeli adolescents. Foods 11, 2072.CrossRefGoogle ScholarPubMed
Hebestreit, A, Barba, G, De Henauw, S, et al. (2016) Cross-sectional and longitudinal associations between energy intake and BMI z-score in European children. Int J Behav Nutr Phys Act 13, 23.CrossRefGoogle ScholarPubMed
Huang, TT, Howarth, NC, Lin, BH, et al. (2004) Energy intake and meal portions: associations with BMI percentile in U.S. children. Obes Res 12, 18751885.CrossRefGoogle ScholarPubMed
Okubo, H & Sasaki, S (2007) Underreporting of energy intake among Japanese women aged 18–20 years and its association with reported nutrient and food group intakes. Public Health Nutr 7, 911917.CrossRefGoogle Scholar
Millstein, SG, Petersen, AC & Nightingale, EO (1994) Adolescent health promotion: rationale, goals, and objectives. In Promoting the Health of Adolescents: New Directions for the Twenty-First Century, pp. 424 [Millstein, SG, Petersen, AC & Nightingale, EO, editors]. Oxford: Oxford University Press.Google Scholar
Farooq, A, Martin, A, Janssen, X, et al. (2020) Longitudinal changes in moderate-to-vigorous-intensity physical activity in children and adolescents: a systematic review and meta-analysis. Obes Rev 21, e12953.CrossRefGoogle ScholarPubMed
Kemp, BJ, Cliff, DP, Chong, KH, et al. (2019) Longitudinal changes in domains of physical activity during childhood and adolescence: a systematic review. J Sci Med Sport 22, 695701.CrossRefGoogle ScholarPubMed
Biddle, SJ, García Bengoechea, E & Wiesner, G (2017) Sedentary behaviour and adiposity in youth: a systematic review of reviews and analysis of causality. Int J Behav Nutr Phys Act 14, 43.CrossRefGoogle ScholarPubMed
Skinner, AC, Steiner, MJ & Perrin, EM (2012) Self-reported energy intake by age in overweight and healthy-weight children in NHANES, 2001–2008. Pediatric 130, e936942.CrossRefGoogle ScholarPubMed
Movassagh, EZ, Baxter-Jones, ADG, Kontulainen, S, et al. (2017) Tracking dietary patterns over 20 years from childhood through adolescence into young adulthood: the Saskatchewan pediatric bone mineral accrual study. Nutrients 9, 990.CrossRefGoogle ScholarPubMed
Schneider, BC, Dumith, SC, Araújo de França, GV, et al. (2018) Tracking and changes in dietary patterns of adolescents: a longitudinal study in Southern Brazil. Nutr Food Sci 8, 16.CrossRefGoogle Scholar
Figure 0

Table 1. Participants’ physical characteristics as well as energy and macronutrient intakes by sex in 2004 and 2009

Figure 1

Table 2. Tracking of energy and macronutrient intakes between baseline and the fifth follow-up year as estimated by correlation coefficients using a mixed effects model and weighted kappa statistics (k)

Figure 2

Table 3. Mixed effects models between total energy intake, macronutrient intakes and time spent in MVPA, sedentary time, sex and age with adolescent BMI over 5 years

Figure 3

Table 4. Mixed effects models between total energy intake, percentage energy from protein, fat and CHO and time spent in MVPA, sedentary time, sex and age with adolescent BMI over 5 years