Introduction

Physical growth is an important aspect of children’s health. Growth retardation used to be a problem in developing countries including Vietnam1. However, in recent years, overweight and obesity has become a common health issue of children in Vietnam2,3,4,5,6,7, many countries in Asia8 and globally9. There have been some reports in some other countries regarding the effects of overweight and obesity on growth and other health issues of children and adolescents10,11 as well as how successful obesity treatment might normalize children growth trajectory and health status12,13,14.

In Vietnam, there have been a few studies evaluating children physical growth1,15 and a few studies reporting overweight and obesity in children2,3,4,5,6,7 but most of these studies had small sample size and was conducted in small regions. There is still limited published data regarding physical growth of Vietnamese children as compared to international standards such as World Health Organization (WHO) or Center of Disease Control (CDC). There is still a lack of large population-based study surveilling growth and especially evaluating overweight and obesity of a large number of children and adolescents across many regions and many age groups over many years to examine population secular trends and regional disparities. There is also a lack of study evaluating the effects of overweight and obesity on height growth of children and adolescents. Such large studies would be essential to understand the specific growth patterns and overweight, obesity issues of Vietnamese children and adolescents at population level.

In addition, growth of children may be different between regions and ethnicities16. WHO reference growth charts may not necessarily be applicable to all countries. As such building growth references based on domestic data would be essential for specific countries like Vietnam. However, to date, there has not been growth chart references specifically for Vietnamese children which were constructed based on large Vietnamese children data.

To address the aforementioned gap and needs, our study analyzed height, weight, body mass index (BMI) for age and weight for height data from annual health check from 2018 to 2024 of children aged from 18 months to 18 years attending private nursery to high schools in 3 major cities in Vietnam including Hochiminh city, Hanoi and Haiphong. The aims were to evaluate physical growth, temporal trends and regional disparities in physical growth as well as issues regarding physical growth especially overweight and obesity and the effects of overweight and obesity on height growth Vietnamese children and adolescents. Our study data was also used to build growth chart references for Vietnamese children.

Results

The following analyses and results will be presented in order: (1) Data summary by city, age, gender, calendar years; (2) Overall data examination regarding weight-for-height, BMI, weight, height for age by gender and comparison with WHO 2007 references to provide an overview of these anthropometrics indices in all study data and in comparison with WHO reference; (3) Describing how these indices BMI, height, weight for age by gender differed between cities (regional disparities); (4) Describing how these indices BMI, height, weight for age by gender changed over calendar years (temporal trends); (5) Describing the prevalence of nutritional status classification according to WHO references: overweight and obesity, wasting or thinness, underweight, stunting to examine the level of public health significance of these nutritional issues in overall and examine the differences between cities and the change over calendar years of these nutritional issues; (6) Describing how these nutritional status classification especially overweight and obesity at prepuberty age affected later height growth; (7) Presenting the growth charts for children using study data as references (may be used for Vietnam until better reference growth charts based on more representative data of Vietnamese children are available).

Data summary

There was a total of 203,550 annual health check visit records of children attending private schools in Hochiminh city, Hanoi and Haiphong from 2018 to 2024. Of which, 1387 visit records (0.7%) were excluded due to missing or erroneous data of the key variables. The remaining 202,163 health check visit records of 88,884 children were included in the analysis.

There were 50,648 (25%) visits from Hochiminh City, 136,650 (67.6%) visits from Hanoi and 14,865 (7.4%) visits from Haiphong. There were 105,298 (52.1%) visits of males and 96,865 (47.9%) visits of females. There were 45,832 (22.7%) visits of children aged from 18 months to 5 years, 98,181 (48.6%) visits of children aged from 5 years to 11 years and 58,150 (28.8%) visits of children aged from 11 years to 18 years. There were 6915 (3.4%) visits in the year 2018, 28,107 (13.9%) visits in the year 2019, 31,294 (15.5%) visits in the year 2020, 46,882 (23.2%) visits in the year 2022, 43,875 (21.7%) visits in the year 2023 and 45,090 (22.3%) visits in the year 2024. The proportion of males and females was similar in Hochiminh city (South of Vietnam) (49.9% of males) while the proportion of males was slightly higher in two cities in the North of Vietnam (Hanoi (52.7% of males) and Haiphong (53.4% of males)). The proportion of age categories and proportion of visits by calendar years were slightly different between the 3 cities (Table 1).

Table 1 Data summary.
Full size table

Weight-for-height, BMI, weight, height for age by gender in overall data and comparison with WHO 2007 references

The median and 95% confidence interval (95%CI) of weight-for-height (height in cm) (Fig. 1a), BMI-for-age (Fig. 1b, Supplementary Fig. 1), height-for-age (Fig. 2), weight-for-age (Supplementary Fig. 2) (age in month) by gender were displayed in comparison with WHO 2007 median and SD references and summarized in Supplementary Table 1.

For age from 18 months to 5 years old, the median weight-for-height and BMI-for-age of males and females was almost the same or slightly higher than WHO reference median. Specifically, the median weight-for-height of males was almost the same as WHO reference median for height < 108 cm and was significantly higher than WHO reference median for height ≥ 108 cm (between the median and + 1SD of WHO reference for height from 108 cm to 114 cm and was between + 1SD and + 2SD of WHO reference for height > 114 cm). The median weight-for-height of females was almost the same as WHO reference median for height < 110 cm and was between the median and + 1SD of WHO reference for height ≥ 110 cm (Fig. 1a). The median BMI-for-age and weight-for-age of males was almost the same as WHO reference median for age from 18 months to 4 years old. For age from 4 years to 5 years old, the median BMI-for-age and weight-for-age of males was significantly higher than WHO reference median but still between the median and + 1 SD of WHO reference. The median BMI-for-age and weight-for-age of females was almost the same as WHO reference median (Supplementary Fig. 1, Supplementary Fig. 2). The median height-for-age of males and females was almost the same as WHO reference median (Fig. 2a).

Fig. 1
Fig. 1
Full size image

Weight for height and BMI for age compared with WHO 2007 references. Median and 95% confidence interval (95%CI) of (a) weight-for-height (rounded by cm) of males and females from 18 months to 5 years old. (b) body mass index (BMI)-for-age (rounded by month) of males and females from 5 years to 18 years old. Study data (Vschool in red) were plotted against World Health Organization (WHO) 2007 reference median (in green) and standard deviation (SD) (in purple for minus and in blue for plus). Black dots were study individual data.

For age from 5 years to 18 years old, the median BMI-for-age and weight-for-age of males and females was significantly higher than WHO reference median but the magnitude was much larger for males. Specifically, the median BMI-for-age of males was between the median and + 1SD of WHO reference for age from 5 years to 8 years, between + 1 SD to + 2 SD of WHO reference for age between 8 years to 13 years and between the median and + 1 SD of WHO reference for age from 13 years to 18 years. The median BMI-for-age of females was slightly significantly higher than WHO reference median (between the median and + 1 SD of WHO reference) for age between 8 years and 13 years old, and was almost the same as WHO reference median for age from 5 years to 8 years and 13 years to 16 years and was between the median and − 1SD of WHO reference for age from 16 years to 18 years old (Fig. 1b). The median weight-for-age of males was largely significantly higher than WHO reference median (between + 1 SD to + 2 SD of WHO reference for age from 8 years to 10 years). The median weight-for-age of females was slightly higher than WHO reference (between the median and + 1 SD of WHO reference for age from 8 years to 10 years) (Supplementary Fig. 2).

For age from 5 years to 18 years old, the median height-for-age of males and females was slightly higher than WHO reference median at prepuberty to early puberty age but became lower than WHO reference median at the end of puberty age. Specifically, the median height-for-age of males was slightly significantly higher than that of WHO reference median (between the median and + 1 SD of WHO reference) before 15 years old, was equal to WHO reference median around the age of 15 years and was slightly significantly lower than that of WHO reference median (between the median and − 1 SD of WHO reference) from 16 to 18 years old. At 18 years old, the median (95%CI) height of males of our data (172 (171.5, 173) cm) was 4 cm shorter than WHO reference median height of males (176 cm). The median height-for-age of females followed a similar pattern to that of males but the crossing point between the median curve of our data vs. WHO reference median curve was around 13 years old. Specifically, the median height-for-age of females was slightly significantly higher than that of WHO reference median (between the median and + 1 SD of WHO reference) before 13 years old, equal to WHO reference median around 3 years old and was slightly significantly lower than that of WHO reference median (between the median and − 1 SD of WHO reference) from 14 to 18 years old. At 18 years old, the median (95%CI) height of females of our data (160 (159, 160) cm) was 3 cm shorter than WHO reference median height of females (163 cm) (Fig. 2, Supplementary Table 1).

Fig. 2
Fig. 2
Full size image

Height for age compared with WHO 2007 references. Median and 95% confidence interval (95%CI) of height-for-age (rounded by month) of study data (Vschool in red) were plotted against World Health Organization (WHO) 2007 reference median (in green) and standard deviation (SD) (in purple for minus and in blue for plus) for (a) males and females from 18 months to 5 years old. (b) males and females from 5 years to 18 years old. Black dots were study individual data.

Regional disparities in BMI, height, weight for age by gender

The median (95%CI) of height, weight, BMI for age by gender for each of the three cities (Hochiminh city, Hanoi and Haiphong) was displayed in Fig. 3. The difference in the medians (95%CI of the difference) of height, weight, BMI for age by gender between Hochiminh city vs. Hanoi and between Hochiminh city vs. Haiphong was summarized in Supplementary Tables 2 and Supplementary Table 3 respectively. The median height of males and females in Hochiminh city was slightly significantly taller than that in Hanoi and Haiphong across most age in years. The median height of males and females in Haiphong was shortest for most ages as compared to Hanoi and Hochiminh city. For example, at 8 years old, the difference in median height (95%CI) of males in Hochiminh city vs. Hanoi and of females in Hochiminh city vs. Hanoi was 1 (1, 1.7) and 0.5 (0, 0.7) cm respectively. At 8 years old, the difference in median height (95%CI) of males in Hochiminh city vs. Haiphong and of females in Hochiminh city vs. Haiphong was 2 (2, 3.5) and 1 (0.95, 2) cm respectively. At 17 years old, the difference in median height (95%CI) of males in Hochiminh city vs. Hanoi and of females in Hochiminh city vs. Hanoi was 1.35 (0.7, 2.3) and 1 (0, 1.3) respectively. At 17 years old, the difference in median height (95%CI) of males in Hochiminh city vs. Haiphong and of females in Hochiminh city vs. Haiphong was 2.5 (1, 4) and 1 (0, 2) respectively (Fig. 3a, Supplementary Table 2, Supplementary Table 3). The median weight for age in Hochiminh city was significantly bigger than that in Hanoi for males and females for most age in years (Fig. 3b, Supplementary Table 2, Supplementary Table 3). The median BMI in Hochiminh city was significantly bigger than that in Hanoi for males and females from 7 to 13 years old. There was no significant difference in median weight and median BMI between Hochiminh city and Haiphong for males and females for most age in years. For both males and females from 11 years to 15 years old, the median BMI of Haiphong was biggest as compared to Hanoi and Hochiminh city (Fig. 3c, Supplementary Table 2, Supplementary Table 3).

Fig. 3
Fig. 3
Full size image

Height, weight, and BMI for age by gender and cities. Median and 95% confidence interval (95%CI) of (a) height for age (in year from 2 to 18). (b) weight for age (in year from 2 to 18). (c) body mass index (BMI) for age (in year from 2 to 18). The data were plotted by gender for each of the three cities (HCP: Hochiminh, HHN: Hanoi, HHP: Haiphong) to evaluate the difference between cities. Median of height, weight, BMI for age (year) x was calculated by taking the median of height, weight, BMI of age (year) from x-0.5 to x+0.5, respectively (for example, median height for 8 years old was calculated by taking the median height for age from 7.5 to 8.5 years old). Confidence interval of median was calculated using nonparametric bootstrap percentile intervals using the “DescTools” package version 0.99.57 and the “boot” package version 1.3-31.

Temporal trends of BMI, height, weight for age by gender

Temporal trend slopes from 2018 to 2024 and 95% confidence intervals (95%CI) of height, weight and BMI for age (in year from 2 to 18) by gender were displayed in Fig. 4. The actual difference in the medians (95%CI of the difference) of height, weight, BMI for age by gender between the year 2024 and the year 2019 were summarized in Supplementary Table 4. The temporal trend slopes from 2018 to 2024 of height were significantly decreased (negative) for most age < 10 years and were significantly increased (positive) for most age > 10 years for both males and females. For example, at 18 years old, the average height of males increased 0.32 cm per year and the average height of females increased 0.35 cm per year from 2018 to 2024 (Fig. 4a). The temporal trend slopes of weight of males followed similar pattern in that the slopes were significantly decreased (negative) for most age < 10 years and was significantly increased (positive) for most age > 10 years. The temporal trend slopes of weight of females were significantly decreased (negative) for most age < 10 years while there was no significant temporal trend observed for weight of females aged > 10 years (Fig. 4b). The temporal trend slopes of BMI of males were significantly decreased (negative) for most age < 10 years and was significantly increased for age from 14 years to 17 years old. The temporal trend slopes of BMI of females were significantly decreased (negative) for most age < 10 years while there was no significant temporal trend observed for BMI of females aged > 10 years old (Fig. 4c).

Fig. 4
Fig. 4
Full size image

Temporal trend from 2018 to 2024 of height, weight, and BMI by age (year) and by gender. Temporal trend slopes and 95% confidence intervals (95%CI) of (a) height. (b) weight. (c) body mass index (BMI). The trend slopes (coefficients) and 95%CI estimated from Generalized Estimating Equations (GEE) models were plotted by age (in year from 2 to 18) and by gender together with observed values and were summarized in the adjacent table. The correlation structure for GEE models was selected based on the smallest Quasi Information Criterion (QIC). For height, weight, BMI of both males and females, “independent” correlation structure had the smallest QIC and thus was selected (detailed results regarding QIC of GEE models with different correlation structures was summarized in Supplementary Table 6). The results of the GEE models with “independent” correlation structure were displayed here. The calendar year from 2018 to 2024 was standardized by minus 2018 (i.e. after standardization, the year 2018 became 0, and the year 2024 became 6 and so on). The GEE models for height, or weight, or BMI contained height, or weight, or BMI as dependent variable, respectively, and standardized calendar year, age in years and interaction term between calendar year and age in years as independent variables. The GEE models were fitted separately for males and females. Result interpretation: The positive slopes with 95%CI which did not include zero were interpreted as significant increase (e.g. the height of males at 18 years old significantly increased by 0.32 cm (95%CI = 0.08 cm to 0.57 cm) per year from 2018 to 2024). The negative slopes with 95%CI which did not include zero were interpreted as significant decrease (e.g. the BMI of males at 8 years old significantly decreased by −0.14 kg/m2 (95%CI = −0.17 kg/m2 to -0.1 kg/m2) per year from 2018 to 2024). The slopes with 95%CI which included zero were not significant (no significant increase nor decrease trend).

Regarding the actual differences of height, weight, BMI of the year 2024 vs. 2019, at 18 years old, the median height of males in 2024 was 2.65 cm taller than that in 2019 (95%CI of median difference was (1.1, 5.5) cm). Similarly, at 18 years old, the median height of females in 2024 was 2.9 cm taller than that in 2019 (95%CI of median difference was (1.3, 7.3) cm) (Supplementary Table 4). The actual differences of height, weight, BMI of the year 2024 vs. 2019 were consistent with the temporal trend slopes estimated from the GEE models above indicating that the GEE models were well fitted the data.

Nutritional status classification according to WHO references

Overweight and obesity

For age from 18 months to 5 years, the prevalence of overweight and obesity of males and females was medium. Specifically, the prevalence (in %) (95%CI) of overweight and obesity of males (5.8% (5.4%, 6.2%) and 3.2% (2.8%, 3.5%) respectively) was slightly higher than that of females (3.9% (3.6%, 4.2%) and 1.2% (0.9%, 1.5%) respectively). The prevalence (95%CI) of overweight and obesity combined was 9.0% (8.6%, 9.3%) for males and was 5.1% (4.8%, 5.4%) for females. The prevalences of overweight and obesity combined for both males and females were regarded as “medium” according to WHO prevalence cut-off values for public health significance (Table 2). There was no clear temporal trend from 2018 to 2024 regarding the prevalences of overweight and obesity of males and females. The prevalence of overweight and obesity in Haiphong was significantly higher than that in Hanoi and Hochiminh city for males (p < 0.0001) and was not clearly different between cities for females (Fig. 5, Supplementary Fig. 3a).

Table 2 Prevalence of overweight, obesity, underweight, wasting or thinness, and stunting by gender and age group.
Full size table

For age from 5 years to 18 years, the prevalence of overweight and obesity of males and females was very high. Specifically, the prevalence (95%CI) of overweight of males (24.8% (24.4%, 25.1%) ) and of females (19.6% (19.3%, 19.9%)) was very high. The prevalence (95%CI) of obesity of males (22.9% (22.5%, 23.3%)) was very high and was much higher than that of females (6.7% (6.4%, 7.1%)). The prevalence (95%CI) of overweight and obesity combined for males was 47.7% (47.3%, 48.0%) and was more than three times larger than the WHO prevalence cutoff value for “very high” public health significance (≥ 15% was classified as very high public health significance according to WHO) (Table 2). Strikingly, for males aged from 8 years to 13 years, the median BMI was between + 1SD and + 2SD of WHO references and thus > 50% was classified as overweight or obesity (Fig. 1b). The prevalence (95%CI) of overweight and obesity combined for females was 26.3% (26.0%, 26.7%) and was nearly two times larger than the WHO prevalence cutoff value for “very high” public health significance (Table 2). There was a significant decrease trend from 2018 to 2024 regarding the prevalence of obesity and overweight of males and the prevalence of overweight of females (p trend < 0.0001). The prevalence of overweight and obesity of both males and females was highest in Haiphong followed by Hochiminh city and Hanoi. The prevalence of overweight and obesity of both males and females in Haiphong and Hochiminh city was significantly higher than that in Hanoi (p < 0.0001) (Fig. 5, Supplementary Fig. 3b).

Fig. 5
Fig. 5
Full size image

Classification of overweight, obesity, wasting or thinness according to WHO 2007 references: temporal trend and difference between cities. Overweight, obesity, and wasting or thinness were classified according to World Health Organization (WHO) recommendation. For children from 18 months to 5 years of age, overweight was weight-for-height (WFH) > + 2 SD above WHO Child Growth Standards median, obesity was weight-for-height > + 3 SD above the WHO Child Growth Standards median, and wasting was weight-for-height <-2 SD of the WHO Child growth standards median. For children aged between 5 and 18 years, overweight was Body Mass Index (BMI)-for-age > + 1 SD above the WHO Growth standards median, obesity was BMI-for-age > + 2 SD above the WHO Growth standards median. Thinness was BMI-for-age <-2 SD of the WHO Child growth standards median. (a) Temporal trend from 2018 to 2024 for males and females from 18 months to 5 years of age. (b) Temporal trend from 2018 to 2024 for males and females from 5 years to 18 years of age. (c) Difference between the three cites Hochiminh (HCP), Hanoi (HHN), Haiphong (HHP) for males and females from 18 months to 5 years of age. (d) Difference between the three cites Hochiminh (HCP), Hanoi (HHN), Haiphong (HHP) for males and females from 5 years to 18 years of age. *Statistical significance for temporal trend or statistical significance for difference between cities. Cochran-Armitage test for trend in proportion was used to evaluate the trend of height, weight, and BMI categories over calendar year. Chi-squared test was used to test the difference in proportion between cities.

Wasting or thinness, underweight, stunting

The prevalence of wasting or thinness and underweight was low or very low for both males and females from 18 months to 18 years. Specifically, for age from 18 months to 5 years old, the prevalence (95%CI) of wasting and underweight of males (1.4% (1.0%, 1.8%) and 1.6% (0.9%, 2.1%) respectively) and females (1.3% (1.0%, 1.6%) and 1.7% (1.0%, 2.2%) respectively) was very low. For males from 5 years to 18 years, the prevalence (95%CI) of thinness (2.5% (2.1%, 2.8%)) and underweight (1.2% (0.7%, 1.7%)) was very low. For females from 5 years to 18 years, the prevalence (95%CI) of thinness (3.3% (3.0%, 3.6%)) was low and the prevalence (95%CI) of underweight (1.9% (1.2%, 2.2%)) was very low (Table 2). There was no clear temporal trend from 2018 to 2024 and no clear difference in the prevalence of wasting or thinness and underweight of males and females between the 3 cities Hochiminh, Hanoi and Haiphong (Fig. 5, Supplementary Fig. 3, Supplementary Fig. 4, Supplementary Fig. 5, Supplementary Fig. 6).

The prevalence of stunting was low or very low for both males and females from 18 months to 18 years. Specifically, for age from 18 months to 5 years old, the prevalence (95%CI) of stunting of males (3.2% (3.2%, 3.2%)) and females (2.8% (2.8%, 2.8%)) was low (Table 2). There was a significant increase trend in the prevalence stunting of males and females from 2018 to 2024 (p trend < 0.0001). The prevalence of stunting of males and females was significantly higher in Hanoi and Haiphong than that in Hochiminh city (Fig. 6, Supplementary Fig. 7). For age from 5 years to 18 years, the prevalence (95%CI) of stunting was very low for both males (0.7% (0.7%, 0.7%)) and females (1.0% (1.0%, 1.0%)) (Table 2). There was no clear temporal trend over calendar years and no clear difference between cities regarding the prevalence of stunting of males and females aged from 5 years to 18 years (Fig. 6, Supplementary Fig. 7).

The prevalence (95%CI) of more detailed categories of height, weight, BMI for age and weight for height according to WHO SD references was summarized in Supplementary Table 5.

Fig. 6
Fig. 6
Full size image

Classification of height-for-age according to WHO: temporal trend and difference between cities. Height-for-age was classified according to World Health Organization (WHO) references by calendar years from 2018 to 2024 and by cities (HCP= Hochiminh city, HHN= Hanoi, HHP= Haiphong). Stunting was height-for-age <-2 SD of the WHO Child growth standards median. Extreme tallness was height-for-age >+3 SD of the WHO Child growth standards median. (a) Temporal trend from 2018 to 2024 for males and females from 18 months to 5 years of age. (b) Temporal trend from 2018 to 2024 for males and females from 5 years to 18 years of age. (c) Difference between the three cites Hochiminh (HCP), Hanoi (HHN), Haiphong (HHP) for males and females from 18 months to 5 years of age. (d) Difference between the three cites Hochiminh (HCP), Hanoi (HHN), Haiphong (HHP) for males and females from 5 years to 18 years of age. SD: Standard deviation. *Statistical significance for temporal trend or statistical significance for difference between cities. Cochran-Armitage test for trend in proportion was used to evaluate the trend of height, weight, and BMI categories over calendar year. Chi-squared test was used to test the difference in proportion between cities.

The effects of overweight, obesity and height at prepuberty age on height growth

Overweight, obesity and height growth

Children with overweight and obesity had taller height at prepuberty age but their height at the end of puberty was not different from their normal or thin peers. Specifically, for males, at 11 years old, the average height of males was tallest in those with obesity, followed by those with overweight, normal BMI for age and thinness (p < 0.0001). However, the height growth of those with obesity or overweight was slowdown earlier than those with normal BMI or thinness. As such, the average height of males with obesity, or overweight or normal BMI, or thinness at 11 years old when they reached 16 to 17 years old was the same (Fig. 7a). This pattern was also similar in females. At 9 years old, the average height of females was tallest in those with obesity, followed by those with overweight, normal BMI for age and thinness (p < 0.0001). However, the average height of females with obesity, or overweight or normal BMI, or thinness at 9 years old when they reached 14 to 15 years old was the same (Fig. 7b). The effects of BMI at 9 years to 14 years of age on height growth for males and the effects of BMI at 7 years to 12 years of age on height growth for females were displayed in Supplementary Fig. 8 and Supplementary Fig. 9 respectively. The above pattern depicted that early taller height in those with obesity or overweight as compared to normal BMI or thinness peers was not associated with taller height at the end of puberty.

Height classification at prepuberty age and height growth

Taller height classification at prepuberty age was associated with taller height at the end of puberty and vice versa. Specifically, for males, at 11 years old, the difference in height between heigh-for-age categories according to WHO references was large. The difference in height of those with + 3SD < height ≤ + 4SD vs. those with − 3SD ≤ height < -2SD of WHO references was approximately 40 cm. Males with lower height categories at 11 years old remained shorter than their peers with higher height categories at 11 years old when they reached 16 to 17 years old. However, around 16 to 17 years old the difference in height between these categories was smaller (corresponding difference was approximately 25 cm) (Fig. 7c). This pattern was also similar in females. At 9 years old, the difference in height between heigh-for-age categories according to WHO references was large in that the difference in height of those with + 3SD < height ≤ + 4SD vs. those with − 3SD ≤ height < -2SD of WHO references was approximately 35 cm. Females with lower height categories at 9 years old remained shorter than their peers with higher height categories at 9 years old when they reached 14 to 15 years old. However, around 14 to 15 years old the difference in height between these categories was smaller (corresponding difference was approximately 15 cm) (Fig. 7d). The above pattern depicted that taller height at prepuberty age was associated with taller height at the end of puberty and vice versa.

Fig. 7
Fig. 7
Full size image

Effects of BMI and height at prepuberty age on height growth. LOESS smooth curve and shaded 95% confidence interval of height for age. (a) By body mass index (BMI) categories at 11 years of age for males. P-values at 11 years and 17 years of age for the difference of height between BMI categories at 11 years of age were shown. (b) By BMI categories at 9 years of age for females. P-values at 9 years and 15 years of age for the difference of height between BMI categories at 9 years of age were shown. BMI was categorized according to World Health Organization (WHO) 2007 references: overweight was BMI-for-age greater than +1 SD above the WHO Growth standards median, obesity was BMI-for-age greater than +2 SD above the WHO Growth standards median. Thinness was BMI-for-age <-2 SD of the WHO Child growth standards median. Normal BMI was BMI-for-age between -2SD and +1SD of WHO Growth standards. SD: Standard deviation. p-values from Kruskal-Wallis’ test. (c) By height categories at 11 years of age for males according to WHO 2007 references. (d) By height categories at 9 years of age for females according to WHO 2007 references.

Reference growth charts for Vietnamese children using study data

Using our data, percentile growth charts at 0.4, 2, 10, 25, 50, 75, 90, 98 and 99.6 percentile (equivalent to −2.68, −2.01, −1.28, −0.67, 0 (mean), + 0.67, + 1.28, + 2.01, + 2.68 standard deviation (SD)) were constructed. Height-for-age growth charts by gender for children from 18 months to 5 years old and for children from 5 years to 18 years old were displayed in Fig. 8. Weight, BMI for age and weight for height growth charts by gender for children from 18 months to 5 years old and for children from 5 years to 18 years old were displayed in Supplementary Fig. 10, Supplementary Fig. 11, and Supplementary Fig. 12.

Fig. 8
Fig. 8
Full size image

Height-for-age growth charts based on study data. Height-for-age percentile growth charts at 0.4, 2, 10, 25, 50, 75, 90, 98 and 99.6 percentile (equivalent to −2.68, −2.01, −1.28, −0.67, 0 (mean), +0.67, +1.28, +2.01, +2.68 standard deviation (SD)) were displayed. Growth charts were constructed using lambda-mu-sigma (LMS) method with Box-Cox t distribution and log link for mu using the “gamlss” package version 5.4-22. (a) Height-for-age percentile growth charts for males and females from 18 months to 5 years of age. (b) Height-for-age percentile growth charts for males and females from 5 years to 18 years of age.

Discussion

We have presented comprehensive analyses and results of one of the largest population-based data regarding physical growth of Vietnamese children and adolescents. Overweight and obesity was strikingly epidemic in our study population, especially in adolescents with prevalence more than three folds higher than WHO cutoff values for very high public health significance for males and nearly two folds higher than WHO cutoff values for very high public health significance for females from 5 years to 18 years of age19,20. Even more dramatically, more than half of males aged from 8 years to 13 years was classified as overweight or obesity. This indicates that overweight and obesity is an extremely important health issue of Vietnamese private school adolescents in urban areas.

The prevalence of overweight and obesity of males and females aged from 18 months to 5 years of our data was less striking and was similar to that (7.4%) of a landscape analysis on childhood overweight and obesity for Vietnamese children using 2000–2020 data done by United Nations Children’s Fund (UNICEF)4,7. The prevalence of overweight of males and females aged between 5 years and 18 years of our data was similar to that of the landscape analysis using 2020 data (19%). The prevalence of obesity of females aged between 5 years and 18 years in our data was slightly higher than that of the landscape analysis using 2020 data (6.7% vs. 4.6%). However, the prevalence of obesity of males aged between 5 years and 18 years in our data was much higher than that of the landscape analysis using 2020 data (23% vs. 10.9%). The difference in prevalence of obesity between males and females aged between 5 years and 18 years in our data was also much larger than that of the landscape analysis4,7. The prevalence and the difference in prevalence of obesity in school age males and females in our data was quite similar to some recently published studies in Vietnam2,3,5 and a recent systematic review and meta-analysis of Vietnam published data up to 20226. This indicates that our anthropometric data might not be very different from general urban children population in Vietnam. In the other hand, the prevalence of obesity of adolescents in our data was quite higher than the prevalence of obesity reported in a meta-analysis of Southeast Asian countries (adolescent males 10.1% and females 6.2%)8. Community education regarding proper diet, physical activities and health monitoring is essential for overweight and obesity prevention and control in children and adolescents. For our specific study population of private school children, follow-up health monitoring, education and interventions in addition to mandatory annual health check would be necessary to reduce the burden of overweight and obesity.

Our data showed that, in general, taller height at prepuberty age was associated with taller height at the end of puberty and vice versa. However, our data demonstrated that children with overweight or obesity had taller height than their peers with normal BMI or thinness at prepuberty age but their height at the end of puberty was not different from their normal BMI or thinness peers. The earlier fast height gain of children with overweight or obesity at prepuberty age followed by subsequent earlier height velocity reduction or lesser growth spurt during puberty is probably due to earlier puberty and endocrinological change because of their overweight, obesity status10,11,21,22. As the prevalence of overweight and obesity was very high in our data, overweight and obesity might drive the overall pattern of height-for-age growth of study data as compared to WHO references in that the median height of our data was taller than WHO median height reference at prepuberty to early puberty age and was flattened earlier and reached lower height than WHO median height reference at 18 years old. Successful treatment of obesity for children and adolescents by dietary changes and increased physical activity may normalize the growth velocity pattern and reduce related health risks due to obesity12,13,14.

Regarding temporal trends, the landscape analysis of UNICEF showed an increase in prevalence of overweight and obesity from 2000 to 2010 and to 20204,7,23. Whereas our data showed a statistically significant decrease in the prevalence of overweight and obesity from 2018 to 2024. This decrease might be due to the increase in awareness regarding overweight and obesity of the families of private schoolchildren in our data. Another good temporal trend observed in our data was the significant increase in median height of both males and females at the end of their puberty age from 2018 to 2024. This increase trend in height over time in our data is consistent with the increase trend in height in many other countries such as China24, Japan and South Korea25. This might be attributed to improved nutrition and better health and healthcare. However, the median height of males and females in our 2018 to 2024 data was still 3 to 4 cm shorter than WHO 2007 median reference. This is consistent with the previously reported lower height of Vietnamese school-aged children and adolescents as compared to WHO reference or those in many other countries26 but the difference between our data and WHO reference was smaller. This might be because our data was of private school children and adolescents who might have better height than general Vietnamese children and adolescents.

Regarding the difference between cities, our data showed that the prevalence of overweight, obesity was highest in Haiphong and the median height of both males and females was shortest in Haiphong as compared to Hanoi and Hochiminh city. As such, educational and interventional activities for overweight and obesity prevention and control should be implemented for all three cities with more emphasis on Haiphong.

The prevalence of stunting, underweight, wasting or thinness in our data was low or very low according to WHO cut-off values for public health significance19,20. These prevalences are lower than previously reported prevalences of stunting, wasting or thinness of Vietnamese children in rural area15, in some countries in Southeast Asia and in some other low middle income countries (LMICs)27,28. These results suggest that stunting, wasting or thinness and related malnutrition issues might not be a significant concern in our study population of children of private schools in major cities in Vietnam. This may also suggest that stunting, wasting or thinness and related malnutrition issues have been improved in Vietnamese children in recent years.

It is worth mentioning that there has not been references for height, weight, BMI for age and weight for height specifically for Vietnamese children which was constructed based on large Vietnamese children data. Therefore, the growth charts based on our large population-based data may be used as references for growth evaluation for Vietnamese children until better growth charts based on better and more representative data of Vietnamese children are available. It should be noted that, as the prevalence of overweight and obesity of our data was very high, the BMI and weight for age growth charts based on our data should be used with much caution.

Our annual health check data of private school children and adolescents are among a few population-based electronic databases of schoolchildren health data in Vietnam usable for analysis and evaluation although there are still limitations with our data. Although annual health check is mandatory for all schoolchildren by Vietnamese law, these health check data are mostly poorly recorded and mostly on paper and mostly not usable for analysis or public health purpose. Therefore, there is a need for digitalization, storage and centralized management of schoolchildren annual health check data in Vietnam. These databases when available and well managed would be invaluable for public health evaluation and intervention.

There are some limitations of our study. First, we only have access to the data of schoolchildren of a private school system in 3 majors cities of Vietnam and thus, our data may not be representative for general schoolchildren in Vietnam. Instead, our data may be more representative for schoolchildren in urban regions of Vietnam. In addition, our data are of children attending private schools whose families may be in somewhat higher social-economical class than average Vietnamese children. Dietary patterns of these private schoolchildren may also be somewhat different from general Vietnamese children or children in rural areas or children with lower socioeconomic status. Therefore, the prevalence of malnutrition issues such as underweight, wasting or thinness and stunting may be lower while the prevalence of overweight and obesity may be higher than that of general Vietnamese children. Second, the follow-up time of our data was not long (from 2018 to 2024) and thus it is not possible to observe long term change nor individual longitudinal physical growth trajectories from young children age to adulthood. This limitation may be solved by accumulating more years of data in future analysis.

In conclusion, our annual health check data of private school children and adolescents in three major cities in Vietnam showed that overweight and obesity was an alarming issue especially in adolescent males. There was a significant decrease trend in the prevalence of overweight and obesity and a significant increase trend in median height of children and adolescents from 2018 to 2024. The prevalence of overweight, obesity was highest in Haiphong and the median height of both males and females was shortest in Haiphong as compared to Hanoi and Hochiminh city. Our findings indicate that prevention and control measures should be implemented to reduce the burden and related health issues of overweight and obesity in Vietnamese children and adolescents.

Methods

Study setting and data source

Vinschool is the largest private school system in Vietnam with > 45,000 students from 18 months to 19 years old and > 50 schools from nursery to high school located mostly in major cities including Hanoi, Hochiminh, and Haiphong. A large percentage of Vinschool students stay with Vinschool from nursery to the end of their high school. Annual health check is mandatory for all Vinschool students. Vinmec is the largest private healthcare system in Vietnam. Vinmec provides annual health check for Vinschool students and the health check data of all Vinschool students are stored in Vinmec Health Check (ViHC) data platform.

For this study, we performed retrospective analysis of annual health check data from 2018 to 2024 of all students aged from 18 months to 18 years attending Vinschool from nursery to high schools in 3 major cities in Vietnam including Hanoi, Hochiminh city, and Hai Phong. The data was extracted from Vinmec electronic (ViHC) database. Data of the year 2021 was not available due to school closure and health check cancelation during COVID-19 outbreak. This database contained data of clinical examination during health check. The extracted data was de-identified and data regarding city, year of health check, gender, age, height, weight and BMI of students recorded at health check were used for the analysis of this report. These data were available for all Vinschool students except some records with data errors which were excluded. This study was approved by Vinmec Ethical Committee (approval number 0231/2024/CN/HDDD VMEC). Informed consent was waived by Vinmec Ethical Committee for retrospective analysis and report of de-identified data. The study, all methods and report were carried out in accordance with the Declaration of Helsinki and STROBE (STrengthening the Reporting of OBservational studies in Epidemiology) guideline. STROBE checklist with our study corresponding status was included in Supplementary Note 1 of Supplementary Information.

Nutritional status classification and comparison with WHO 2007 references

Height, weight, and BMI for age and weight for height were compared with World Health Organization (WHO) Z-score references in 2007 18,29. WHO height and BMI for age reference data was available for comparison with our data from 18 months to 18 years old. WHO weight for age reference data was available for comparison with our data up to 120 months (10 years) old. WHO weight for age reference data from 11 to 18 years old was not available. Classification of height, weight, and BMI by WHO reference standard deviation (SD) by gender was done for age from 18 months to 5 years and for age between 5 years to 18 years overall and by cities and by calendar years.

Overweight, obesity, wasting or thinness, underweight, and stunting were classified according to WHO recommendation17,18. For children from 18 months to 5 years of age, overweight was weight-for-height > + 2 SD above WHO child growth standards median, obesity was weight-for-height > + 3 SD above the WHO child growth standards median, stunting was height-for-age < -2 SD of the WHO child growth standards median, wasting was weight-for-height < -2 SD of the WHO child growth standards median and underweight was weight-for-age < -2 standard deviations (SD) of the WHO child growth standards median19,20. For children aged between 5 and 18 years, overweight was BMI-for-age > + 1 SD above the WHO growth reference median, obesity was > + 2 SD above the WHO growth reference median. A BMI > + 3 SD above the WHO child growth standards median was referred to as severe obesity. Moderate stunting was height-for-age < -2 SD and severe stunting was < -3 SD of the WHO child growth standards median. Moderate underweight was weight-for-age < -2 SD and severe underweight was weight-for-age < -3 SD of the WHO child growth standards median. Moderate thinness was BMI-for-age < -2 SD and severe thinness was BMI-for-age < -3 SD of the WHO child growth standards median.

WHO prevalence cut-off values for public health significance were: (a) For thinness, overweight & obesity: <2.5% = very low, 2.5 to < 5% = low, 5 to < 10% = medium, 10 to < 15% = high, ≥ 15% = very high; (b) For stunting: <2.5% = very low, 2.5 to < 10% = low, 10 to < 20% = medium, 20 to < 30% = high, ≥ 30% = very high.

Statistical analysis

All analysis was done in R statistical software version 4.4.3 30. The median of height, weight and body mass index (BMI) for age were calculated based on the actual age at health check visits. Height, weight, BMI for month or year x was calculated by taking the median height, weight, BMI of month or year from x-0.5 unit to x + 0.5 unit, respectively (for example, height for 8 years old was calculated by taking the median height for age from 7.5 to 8.5 years old). Confidence interval of median was calculated using nonparametric bootstrap percentile intervals31,32,33,34 using the “DescTools” package version 0.99.57 35 and the “boot” package version 1.3–31 36,37.

Temporal trend from 2018 to 2024 of height, weight, BMI by age and gender was evaluated by Generalized Estimating Equations (GEE) models using the “geepack” package version 1.3.12 38. The GEE models allowed the estimation of the average trend slopes while taking into account the correlation of repeated measures of the same subjects. The calendar year from 2018 to 2024 was standardized by minus 2018 (i.e. after standardization, the year 2018 became 0, and the year 2024 became 6 and so on). The GEE models for height, or weight, or BMI contained height, or weight, or BMI as dependent variable, respectively, and standardized calendar year, age in years and interaction term between calendar year and age in years as independent variables. The GEE models were fitted separately for males and females. The correlation structure for GEE models was selected based on the smallest Quasi Information Criterion (QIC). Detailed results regarding QIC of GEE models with different correlation structures were summarized in Supplementary Table 6. For height, weight, BMI of both males and females, “independent” correlation structure had the smallest QIC and thus was selected and thus the results of the GEE models with “independent” correlation structure were presented in Fig. 3. The trend slopes (coefficients) and 95%CI estimated from GEE models were plotted by age (in year from 2 to 18) and by gender together with observed values and were summarized in the adjacent table. The positive slopes with 95%CI which did not include zero were interpreted as significant increase. The negative slopes with 95%CI which did not include zero were interpreted as significant decrease. The slopes with 95%CI which included zero were not significant (no significant increase nor decrease trend).

The difference of height, weight and BMI by age and gender at the first and the last year (2024) of the study period was calculated to evaluate the actual change over time. The year 2019 was chosen as the first year as in the year 2018 there were much fewer children with available data than the following years. The difference in the median of height, weight and BMI by age and gender between the 3 cities was also calculated. Two-sided 95% confidence interval for the value of median(x)-median(y) was calculated based on 999 bootstrap replications and the bias-corrected and accelerated (BCa) method31,32,33,34 using the “confintr” package version 1.0.2 39.

Glaz & Sison’s method for simultaneous confidence intervals for multinomial proportions40 was used to estimate the confidence intervals of the prevalence of height, weight, BMI categories classified according to WHO recommendation. Cochran-Armitage test for trend in proportion was used to evaluate the trend of height, weight, and BMI categories over calendar year. Chi-squared test was used to test the difference in proportion between cities.

Using our data, we constructed height, weight, BMI for age and weight for height growth charts for Vietnamese children using lambda-mu-sigma (LMS) method41,42 with Box-Cox t distribution and log link for mu 43 using the “gamlss” package version 5.4–22 44,45. Height, weight, BMI for age and weight for height growth charts were constructed by gender for age from 18 months to 5 years. Height, weight, BMI for age growth charts were constructed by gender for age between 5 years to 18 years. Percentile growth charts at 0.4, 2, 10, 25, 50, 75, 90, 98 and 99.6 percentile (equivalent to −2.68, −2.01, −1.28, −0.67, 0 (mean), + 0.67, + 1.28, + 2.01, + 2.68 standard deviation (SD)) were displayed.