Introduction
Body shape plays an important role in many fields. Understanding body shape helps us choose appropriate clothing, design suitable products, evaluate health status, and improve athletic performance. In recent years, Vietnamese people’s body shape has undergone significant changes due to the influence of many factors such as nutrition, lifestyle, environment, etc. Nonetheless, information about Vietnamese male body shape is still limited and not fully updated. This study aim provide scientific information to researchers, designers, and manufacturers based on analysis using effective analytical tools and contribute to the development of scientific information. Developing Vietnam’s fashion, design, pharmaceutical, and sports industries while raising Vietnamese people’s awareness of the importance of health and body shape. In the field of apparel design and manufacturing, understanding variations in body shape plays an important role in ensuring fit, comfort, and aesthetics for consumers (1). Previously, researchers have used many methods to analyze the sophistication of human body morphology. In Vietnam, a number of studies have focused on studying female body shapes from many different aspects. Hien and her colleagues (2) studied the correlation coefficient using 3D V-Stitcher to design basic body blocks for different phenotypes of Vietnamese women. Another study by Hien (3) used a fuzzy logic system to classify female body shapes. In addition, research on male body shape is also of interest in Vietnam, with Kieu and colleagues (4, 5) delving into body classification in Ho Chi Minh City. Meanwhile, Hien and her colleagues (6) developed a system for measuring body size. Furthermore, Nguyen et al. (7) also developed a size system data system for middle-aged men. These studies have contributed valuable insights into the diverse body shapes of Vietnamese people, expanding the scope of body shape analysis to other Asian populations. In China there are also some related studies, such as: Huang and Can used software to analyze and present 4 types of young women’s body shapes (8); Ya-Li Ling & An-Hua Zhong used 3D body measurement technology to captures 5 types of middle-aged women’s shapes with the purpose of improving the fit of clothes (9); and a research article on the method of choosing the size of ready-made clothes using Using fuzzy techniques through the Sugeno model for the MISO fuzzy system (10, 11), there is also a method to build a 3D parametric model according to the lower part of a young woman’s body, which is an important part in the numerical process. chemistry of the garment industry (12). These studies enrich the understanding of Chinese body shape changes. Similarly, research in Korea has shed light on adult body shape (13–15), individuals have large bodies (16), and athletes (17) in Japan. An effective research method for predicting human body type using artificial neural networks and discriminant analysis (18). Additionally, researchers have investigated the possibility of using a virtual clothing system to create 3D avatars reconstructed from 3D human body scan shape data on CLO 3D software (19). They used a variety of methods, including factor analysis, cluster analysis, and body scanning, to identify body types with increasing accuracy.
Therefore, in this study, we will use relevant analyzes using SPSS software to identify the most suitable body shape clustering options and characterize them using CLO3D software. These avatars will provide a more visual presentation and allow for the creation of a database of Vietnamese male body shape avatars.
Methodology
Subjects
The authors randomly measured the first 30 samples to calculate the minimum number of samples that need to be collected. To ensure a 95% confidence level, that is, α = 0.05 with a corresponding t-value of 1.96. From the statistical analysis data of the first 30 samples, the highest standard deviation was determined to be S = 6.74. We also take into account the measurement error of m = 1 cm.
Formula to calculate the number of samples needed for a research article (2):
The results suggest that it is necessary to collect body measurements from at least 175 individuals to ensure the desired level of accuracy and reliability. The authors collected 180 data samples, ensuring the minimum number of samples, and began analyzing. In this study, individuals were randomly selected from many different locations such as companies, parks, and supermarkets. Our data shows that men between the ages of 25 and 35 make up 66.78%, while those between 35 and 45 make up 33.22%. To create the most detailed and accurate avatar in CLO3D software, we meticulously measured each subject to collect data on 30 important body measurements. These dimensions include 10 height measurements, 5 length measurements, 12 circumference measurements, 2 width measurements, and a weight indicator.
Statistical analyses
By utilizing SPSS 27.0 software, the process of classifying body shapes becomes more efficient and accurate. The authoring team employed specific analytical methods, including:
– Principal Component Analysis (PCA), a statistical procedure that transforms highly correlated variables into a smaller number of principal components that capture the majority of the observed variable variance (20).
– K-means clustering to analyze subjects into clusters, where each cluster represents a different body shape type (21).
– Discriminant analysis to provide a more intuitive understanding of the differentiation between body shape clusters identified by K-means clustering.
– Analysis of Variance (ANOVA) to ensure that these clusters represent distinct and statistically significant body types (22).
Following a series of analyses and making selections regarding the number of body shape groups for classification, the team proceeded to use CLO3D to compare representative shapes and ensure their accurate depiction of these body shapes. To set up the avatar’s shape on CLO3D software, 30 measurements of 30 dimensions are required.
Findings
The Varimax rotation method was employed to elucidate the characteristics of the factors. These dimensions all adhere to a normal distribution with high reliability. To ascertain whether the selected variables are suitable for factor analysis, KMO and Bartlett’s test were conducted through an iterative process 6 times to eliminate undesirable variables. Variables eliminated were those with low factor loadings or high factor loadings of 0.5 or more for two or more factors and high loadings for factors with differing concepts (23). After the first EFA analysis results, the author used the method of eliminating bad variables one at a time. From the 30 observed
variables in the first EFA analysis, remove the vertical trunk and underarm and put the remaining 28 observed variables in the second EFA analysis. After the removal of these variables, the results of the KMO sample adequacy test and Bartlett’s test of sphericity were presented in Table 1, showing a KMO value of 0.936 and a significance level (sig.) of 0.000, indicating statistically significant values. Thus, factor analysis is appropriate.
Table 1. KMO coefficients and Bartlett’s Test in the results of the second EFA.
In Table 2, there are 4 factors extracted based on the criterion Eigenvalue > 1, so these 4 factors best summarize the information of the 28 observed variables included in the EFA. The total variance extracted by these 4 factors is 80.84% > 50%, thus, the 4 extracted factors explain 80.84% of the data variation of the 28 observed variables participating in EFA.
Table 2. Total variance explained table.
The rotated matrix results indicate that the 28 observed variables were categorized into 4 factors, with all variables showing factor loadings greater than 0.5 and no poor-quality variables remaining. Using the Varimax rotation method, convergence was achieved in 6 iterations (Table 3), revealing height, waist circumference, thigh circumference, and wrist circumference as the 4 main components.
Table 3. Rotation matrix table in the second EFA results.
Main Component 1: Parameters indicating the body’s vertical dimensions. The highest value is displayed at a height of 0.97, with parameters from thigh length to the length from buttocks to heel all having indices above 0.90, and all other dimensions yielding high results greater than 0.80. This factor indicates an eigenvalue of 14.70 and an accumulated variance ratio of 52.48%. This group of factors is characteristic of the body’s height, and the ratio of dimensions within this factor allows for the determination of the body’s upper and lower proportions.
Main Component 2: Parameters indicating the body’s horizontal dimensions. The highest value is displayed at waist circumference at 0.92, followed by hip circumference with 0.92, and parameters from hip circumference to chest circumference all having indices above 0.80, with all other dimensions yielding high results greater than 0.50. This factor shows an eigenvalue of 6.09 and an accumulated variance ratio of 21.75%. This factor may represent the level of obesity in the upper part of the body.
Main Component 3: Parameters indicating the body’s horizontal dimensions. The indices for thigh circumference, calf circumference, and knee circumference are 0.80, 0.799, 0.795, and 0.789, respectively. This factor indicates an eigenvalue of 1.36 and an accumulated variance ratio of 4.86%. This group of components may represent the level of obesity in the lower part of the body.
Main Component 4: The wrist circumference dimension with an index of 0.94. This factor indicates an eigenvalue of 1.05 and an accumulated variance ratio of 3.75%.
Analyzing correlations within the factor matrix
Based on Table 4, “Height” and “Waist Circumference” have a weak positive correlation (0.278**), meaning as height increases, waist circumference may also tend to increase. “Waist Circumference” and “Thigh Circumference” have a moderate positive correlation (0.487**), suggesting that an increase in waist size often accompanies an increase in thigh size. Other correlations are lower and may not be statistically significant, such as “Height” and “Wrist Circumference” at just 0.031, nearly no correlation.
Table 4. Correlation analysis table to evaluate the relationships between variables.
Body shape classification
To analyze body shapes, the research team utilized K-means clustering and discriminant analysis to identify differences among the study groups. Following this, they conducted ANOVA tests on the mean values of independent sample groups and used these test results for body shape classification. In this study, the team proposed 7 solutions with group numbers ranging from 3 to 7 to determine the most suitable grouping. They then performed scatter plot analysis to examine the overlap between the 7 solutions.
After implementing 7 solutions for the body shape analysis, the authors found Solution 3 (Figure 1), which divides individuals into 5 distinct body shape groups, to be the most logical. Solution 3 was chosen because there was no overlap or interweaving between the groups. This showed that there was a difference between the body shape groups in this solution. This is one of the reasons for choosing Solution 3 for body shape analysis. In addition, the ANOVA test results (Table 5) reveal that 30 out of 30 parameters have a significance value (sig) of less than 0.05, indicating significant differences between groups for the variables mentioned in the table (11). The table detailing the dimensions of each body shape group in Table 6. (the table for the 5 groups’ dimensions) shows that individuals within the same group share similar dimensions, and there are noticeable differences between groups. According to Table 7, there’s a significant variation in the number of elements within each group. Although Figure 1c displays some overlap among elements of different groups, it is not substantial. The scatter plot indicates that the elements within each group are not spread too widely, suggesting a relatively cohesive grouping.
Figure 1. (a) Solution 1 with 3 groups; (b) Solution 2 with 4 groups; (c) Solution 3 with 5 groups; (d) Solution 4 with 6 groups; (e) Solution 5 with 7 groups.
Table 5. The ANOVA test table for Solution 3 (27).
Table 6. The dimension parameters table of Solution 3.
Table 7. The proportion of people in the proposed solutions (total 175 people).
Setting up parameters for an Avatar in CLO3D
Having successfully established the parameters and distinguished the 5 body shapes, we were presented with visuals providing an intuitive and comprehensive understanding of each type, showcased through front, back, side, and skew views (Figure 2). A comparison of avatar display results on CLO3D software between the 5 body groups in this study is shown in Figure 3, highlighting the differences in body proportions according to height and circumference among the 5 groups. Armed with these visualizations, our analysis will delve deeper into characterizing each body type through BMI, WHR, and DROP and providing insights tailored to individuals within these body groups.
Figure 2. Classification of 5 body shapes of Vietnamese men aged 25–45 years old (a) front view; (b) back view; (c) side view; (d) skew view.
Figure 3. The sketch in body shapes among 5 different physiques (a) front view (b) side view.
Discussions
Group 1: U-shaped body shape
Individuals in this group have a BMI of 24.9 (kg/m2), which falls into the overweight category (23 ≤ BMI ≤ 24.9) (24). Their waist-to-hip ratio (WHR) of 0.90 indicates a moderate risk of disease according to health evaluations (25). In terms of attractiveness to males, this body type may be less appealing to the opposite sex (26). They are relatively short (shorter than Groups 2, 3, and 4 but taller than Group 5) with horizontally developed bodies. They belong to the medium-short shoulder group with a shoulder drop of 5.31. They also have a slightly hunched back with a backward-leaning posture. The Drop index for Group 1 is 7.25, indicating a smaller waist-to-chest ratio compared to other groups. The height difference between Group 1 and the other groups is relatively large, while the differences in inseam height between Group 1 and the other groups are relatively small, so the legs of Group 1 are longer compared to those of Groups 2, 3, and 4. In short, the authors conclude that this body type resembles a U-shape, which is very common, especially in individuals aged 30 and above. This body type can lead to health imbalances and an increased risk of obesity. It’s important to adopt a healthier diet and engage in regular physical activity to achieve a more balanced physique.
Group 2: Y-shaped body shape
The second group has a body shape with a Y-shaped silhouette. Individuals in this group have a BMI of 21.4 (kg/m2), falling into the normal category (18.5 ≤ BMI ≤ 22.9). Their WHR of 0.87 indicates a low risk of disease according to health evaluations. In terms of attractiveness to males, this body type is somewhat more appealing than the first group. Individuals in Group 2 have average height (shorter than Groups 3 and 4, taller than Groups 1 and 5). They belong to the medium shoulder group with a shoulder drop of 5.07. However, the ratio of inseam height to overall body height is equal to Group 1 and lower than Group 5, so their legs are shorter compared to the body, although slightly taller than Group 1 and 5. The Drop index is 9.17, indicating a larger chest circumference compared to the waist, larger than Group 1 and 5. This means they have relatively broad shoulders and developed chest and shoulders, while the waist and hips are slimmer. In short, the authors conclude that this body type resembles a Y-shape, which is also very common, especially in individuals aged 25 and above.
Group 3: V-shaped body shape
The third group has a body shape with a V-shaped silhouette. Individuals in this group have a BMI of 23.1 (kg/m2), categorizing them as overweight. However, observations of avatar images from the front, back, and side show that this body shape does not exhibit signs of overweight. Instead, it shows well-developed muscles in the chest, shoulders, arms, and legs. This highlights a limitation of BMI, as it fails to differentiate between fat and muscle mass in the body. Their WHR of 0.84 indicates that this body shape is associated with low health risks and is considered the most attractive body shape for males compared to the other 4 groups. They have a tall stature and well-developed upper bodies. This group has a medium shoulder width with a shoulder drop of 5.44. The Drop Index of 17.5 indicates broad shoulders and chest with well-developed muscles in the chest and shoulders, while the waist and hips are slim. Compared to the Y-shaped body of Group 2, Group 3 has broader shoulders (42.71 > 40.25) and a larger chest (98.46 > 89.5), while the waist circumference of the 2 groups is similar (80.96 > 80.33). Overall, the upper body is broader than the lower body. While the difference in height is relatively small (172.82 compared to 172.37), there is a significant variance in inseam height (78.04 > 76.21). Therefore, the legs of Group 3 are relatively longer compared to those of Group 4. In short, the authors concluded that this body shape resembles a V-shape. This body type is quite common, especially among individuals aged 30 and above, and is considered the ideal physique for men. The V-shaped physique enhances male attractiveness, but achieving this requires consistent exercise and proper nutrition.
Group 4: O-shaped body shape
The fourth body shape group has a BMI of 26.16 (kg/m2), which falls into the category of obesity (> 25 kg/m2). Their WHR of 0.93 indicates a high risk of disease, ranking highest among the 5 groups, and this group is considered the least attractive compared to the other 4 groups. This body shape accumulates a significant amount of fat around the abdomen, chest, shoulders, and arms. This group has a low shoulder height measurement of 5.78. The back is slightly hunched compared to group 3. The Drop index of 7.89 indicates that group 4 does not have much difference between chest and waist measurements compared to the other groups. Compared to the U-shaped body of group 1, group 4 has broader shoulders (42.57 > 39.23), a more developed chest (101.95 > 95.83), and relatively different waist-hip measurements. In summary, the authors conclude that this body shape belongs to the O-shaped body type. This body type is quite common, especially in individuals aged 35 and older who are less active, have unhealthy eating habits, and tend to accumulate excess fat around the abdomen.
Group 5: H-shaped body shape
The fifth body shape group has a BMI of 21.3 (kg/m2), falling within the range of 18.5 ≤ BMI ≤ 22.9, categorized as a normal body shape. Their WHR of 0.86 indicates a low risk of disease, and it is somewhat more appealing than Groups 1 and 4. The individuals in Group 5 have a relatively short height (the shortest among the 5 groups) and a slender body compared to the other groups. This group has high shoulders and a shoulder drop of 4.62. The Drop index of 8.47 indicates that Group 5 does not have a significant difference between chest and waist measurements, combined with relatively small chest and waist measurements. Group 5 belongs to the body shape with relatively longer legs compared to Group 1, although this group is shorter than Group 2 (159.85 < 162.97), but with a higher hip level. In summary, the authors conclude that this body shape resembles the H-shaped body. This body type is quite common, especially in individuals aged 25 and older. To maintain and enhance the beauty of the H-shaped body, men need to pay attention to regular exercise and maintain a healthy and balanced diet.
Results
The results of the study categorize individuals into 5 body shape groups: U-shaped, Y-shaped, V-shaped, O-shaped, and H-shaped. Each group is characterized by specific measurements such as BMI, WHR, Drop, shoulder width, shoulder drop, and other body proportions. Table 8 shows a summary of the outstanding characteristics of each physique with accompanying body analysis indicators.
Table 8. Summarizes the shape of the human body.
Conclusion
The results of this study highlight the importance of integrating multiple methods to accurately analyze human body shapes accurately. Based on data collected from 175 men aged 25–45, 5 primary body shape types were identified. The implications of this research are significant across various fields, particularly in the realm of fashion. The research provides foundational data for designing clothes that suit the body shapes of Vietnamese men, thereby enhancing the quality and aesthetics of the product. Designers can also segment the market into different customer groups based on individual needs and preferences. Furthermore, customized product lines can be developed for each customer group, optimizing business efficiency and competitiveness.
In addition, the study contributes valuable insights into understanding body shape variations among Vietnamese men concerning age and region. Medically, these data are highly relevant for diagnosing and treating conditions related to overweight and underweight issues, thereby improving men’s health. Understanding the changes in the male body can also aid ergonomists in designing products and work environments suitable for different age groups, ensuring optimal support for the body and preventing potential health risks.
Nevertheless, with a relatively small sample size, the authors acknowledge limitations in analyzing male physique in the Ho Chi Minh City area, Vietnam. To improve that, the author proposes research and analysis based on a larger sample size to improve the accuracy of the research topic. At the same time, compare data on men’s physique with other countries around the world. Last but not least, using direct manual measurements may cause errors in some measurement dimensions, so it is necessary to apply today’s advanced 3D scanning technology to obtain more accurate measurement results in research. after that.
As a result, the CLO3D software application is an essential tool to help achieve results based on research on Vietnamese men’s body shape.
Acknowledgments
We would like to express our deepest gratitude to those who have accompanied and supported our research results. First of all, we would like to thank our advisors and lecturers, especially Dr. Hien provided his help and profound contributions to help us complete this research. Additionally, we would like to thank our colleagues who encouraged and motivated us to continue our research and for the insightful discussions that supported this research. Therefore, we are extremely grateful for the efforts, companionship, and help of all those mentioned in this study.
Author Contributions
Based on the distribution of works, author Nguyen is responsible for collecting and processing data. At the same time, create design drawings and study the deviation between results between body shapes. In addition, author Yen is responsible for analyzing the data and contributing to the drafting based on the research results. With the companionship and help of Dr. Hien through instructions and suggestions in editing the manuscript for publication. Finally, all authors discussed and made decisions for the final version of the manuscript.
References
1. Zakaria N. Chapter 4 - Body shape analysis and identification of key dimensions for apparel sizing systems. In: Gupta D, Zakaria N , editor. Anthropometry, Apparel Sizing and Design (Woodhead Publishing Series in Textiles). UK: Woodhead Publishing (2014). p. 95–119. doi: 10.1533/9780857096890.1.95
2. Nguyen H, Tran K, Luu L. Study on correlation coefficient to design body basic block for various Vietnam women’s somatotypes by 3D V-Stitcher. Sci Technol Dev J. (2016) 19:70–80.
3. Nguyen HTM. A using the fuzzy logic to classify Vietnamese women’s shapes from 6 to 18 years old. Sci Technol Dev J. (2020) 3:352–65.
4. Kieu T, Tung N, Bao P, Van Anh TT. A study on Vietnamese men’s body classification in Ho Chi Minh City. J Sci Technol. (2022) 58.
5. Kieu T, Tung N, Van Anh T. Classification of body types of Vietnamese middle-aged men in Ho Chi Minh City. JST Eng Technol Sustain Dev. (2022) 32:28–36.
6. Hien N, Quan V, Huong B, Hue T, Duong N. Development of a body measurement sizing system for Southern Vietnamese men aged 18 to 25. J Sci Technol Dev. (2018) 1:25–32.
7. Nguyen P, Tram P, Tu P. Research on Body Shape Characteristics, Development of a Sizing System for Middle-Aged Men, and Its Application in Men’s Clothing Design. Bachelor’s Thesis. Ho Chi Minh City: Ho Chi Minh City University of Technology (2023).
8. Huang C. Study on classification of body shape and size grading on young women of Quanzhou district. Adv Mater Res. (2014) 989–994:5319–22. doi: 10.4028/www.scientific.net/AMR.989-994.5319
9. Ling Y, Zhong A. Research on the classification of body type and prototype of middle-aged women based on 3D scanning. J Fib Bioeng Inf. (2020) 13:161–7. doi: 10.3993/jfbim00344
10. Tan L. Analysis of fuzzy classification of women’s status in China. Chin J Popul Sci. (1991) 3:69–73.
11. Liu K, Wang J, Tao X, Zeng X, Bruniaux P, Kamalha E. Fuzzy classification of young women’s lower body based on anthropometric measurement. Int J Indust Ergon. (2016) 55:60–8. doi: 10.1016/j.ergon.2016.07.008.
12. Wang J, Li X, Pan L, Zhang C. Parametric 3D modeling of young women’s lower bodies based on shape classification. Int J Indust Ergon. (2021) 84:103142. doi: 10.1016/j.ergon.2021.103142.
13. Kim O-K. A study on the body shapes of men at the age of 35-49. Fash Text Res J. (2005) 7:301-308.
14. Kim K, Suh M. Classification and characteristics of the body shape for early adolescent boys. Res J Cost Cult. (2005) 13:344–60.
15. Kim E, Nam Y. Analysis of lower body shape of men in their 30s for pants pattern designs-focus on changes in human dimensions and body type classification. J Kor Fash Cost Des Assoc. (2021) 23:133–46. doi: 10.30751/KFCDA.2021.23.2.133
16. Park W, Park S. Body shape analyses of large persons in South Korea. Ergonomics. (2013) 56:692–706. doi: 10.1080/00140139.2012.752529
17. Tsunawake N, Tahara Y, Yukawa K, Katsuura T, Harada H, Kikuchi Y. Classification of body shape of male athletes by factor analysis. Ann Physiol Anthropol. (1994) 13:383–92. doi: 10.2114/ahs1983.13.383.
18. Kim N, Song H, Kim S, Do W. An effective research method to predict human body type using an artificial neural network and a discriminant analysis. Fibers Polym. (2018) 19:1781–9. doi: 10.1007/s12221-018-7901-0
19. Hong E. Usability verification of virtual clothing system for the production of a 3D avatar reproduced from 3D human body scan shape data - focusing on the CLO 3D program. J Kor Fash Cost Des Assoc. (2020) 22:1–13. doi: 10.30751/KFCDA.2020.22.1.1
20. Niu J, He J, Li M, Zhang X. A comparative study on application of data mining technique in human shape clustering: principal component analysis vs. factor analysis. Proceedings of the 2010 5th IEEE Conference on Industrial Electronics and Applications, Taichung. Piscataway, NJ: IEEE (2010). 2014–18.
21. Ran L, Luo H, Zhang X, Hu H, Zhao C, Liu T. Classifications of body size for Chinese females in three areas. In: Goonetilleke R, Karwowski W , editors. Advances in Physical Ergonomics & Human Factors. AHFE (2018). Advances in Intelligent Systems and Computing (Vol. 789). Cham: Springer (2019).
22. Yu M, Kim D. Body shape classification of Korean middle-aged women using 3D anthropometry. Fash Text. (2020) 7:35. doi: 10.1186/s40691-020-00223-8
23. Cha S. Body shape classification for adult male under 170 cm. J Kor Soc Cloth Text. (2021) 45:1–16. doi: 10.5850/jksct.2021.45.1.1
24. Pan W, Yeh W. How to define obesity? Evidence-based multiple action points for public awareness, screening, and treatment: an extension of Asian-Pacific recommendations. Asia Pac J Clin Nutr. (2008) 17: 370–4.
25. Body Measuring Techniques. East Lancashire Hospitals NHS Trust Website. Available online at: https://elht.nhs.uk/services/dietetics/body-measuring-techniques#:~:text=For%20neck%2C%20chest%20%2F%20bust%2C,measurements%20twice%20to%20improve%20accuracy (accessed March 13, 2024).
26. Fan J, Dai W, Liu F, Wu J. Visual perception of male body attractiveness. Proc R Soc B. (2005) 272:272219–26. doi: 10.1098/rspb.2004.2922
27. Trong H, Ngoc C. Research Data Analysis with SPSS (Vol. 1). Ho Chi Minh City: Hong Duc Publisher (2008).
© The Author(s). 2024 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.