- Email: [email protected]
Head and facial anthropometry of mixed-race US Army male soldiers for military design and sizing: A pilot study
ARTICLE IN PRESS
Applied Ergonomics 36 (2005) 379–383 www.elsevier.com/locate/apergo
Technical note
Head and facial anthropometry of mixed-race US Army male soldiers for military design and sizing: A pilot study Miyo Yokota Geo-Centers, Inc., 190 N. Main Street, Natick, MA 01760, USA Received 5 January 2004; accepted 14 January 2005
Abstract In the United States, the biologically admixed population is increasing. Such demographic changes may affect the distribution of anthropometric characteristics, which are incorporated into the design of equipment and clothing for the US Army and other large organizations. The purpose of this study was to examine multivariate craniofacial anthropometric distributions between biologically admixed male populations and single racial groups of Black and White males. Multivariate statistical results suggested that nose breadth and lip length were different between Blacks and Whites. Such differences may be considered for adjustments to respirators and chemical–biological protective masks. However, based on this pilot study, multivariate anthropometric distributions of admixed individuals were within the distributions of single racial groups. Based on the sample reported, sizing and designing for the admixed groups are not necessary if anthropometric distributions of single racial groups comprising admixed groups are known. r 2005 Elsevier Ltd. All rights reserved. Keywords: Head and face anthropometry
1. Introduction Anthropometry has been critical to successfully designing and sizing protective helmets, eyewear, and respirators in the US Army. Compared to early demography based primarily on White males, the anthropometric distribution of the US Army has changed since individuals of diverse ethnic backgrounds began serving in the 1960s (Bradtmiller et al., 1985; Brues, 1945; Gordon, 1996). Such demographic changes may affect the anthropometric distributions of the populations for which military equipment was designed. Current US Army ergonomics are based on the 1988 US Army anthropometric survey (ANSUR) database (Gordon et al., 1989), weighted for primary racial groups and different age proportions in each gender throughout US Army demography. However, despite the fact that the Tel.: +1 508 233 5845; fax: +1 508 233 5298.
E-mail addresses: [email protected], [email protected]. 0003-6870/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.apergo.2005.01.009
number of biologically admixed populations is increasing in the US (Gibson and Lennon, 1999), there have been few related ergonomic studies. This study investigated differences in multivariate head and face anthropometric distributions between biologically admixed male individuals and single racial male groups (Whites or Blacks). The goal of this study was to examine the effect that an increasingly admixed US Army population might have on current and future equipment sizing and design statistics.
2. Materials and methods Craniofacial anthropometry of biracial male soldiers ðN ¼ 26Þ and admixed soldiers with 14 White ancestry and 34 Black ðN ¼ 8Þ were compared to those of White soldiers ðN ¼ 820Þ and Black soldiers ðN ¼ 1189Þ: Race was determined during interviews with an anthropologist by a subject’s report of ancestry and his parents’ ancestries. The ANSUR survey contains separate
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
380
questions on race and ethnicity because people are more familiar with the term ‘‘race’’ to describe their biological background. All subjects were born in the US. Their ages ranged between 18 and 35 years, and were grouped into 5-year intervals (below 20, 21–25, 26–30, 31–35). The age distribution between White, Black, and mixed groups was homogenous ðp40:05Þ; based on a Fisher’s exact test. Table 1 shows the summary of age distribution by racial groups. Anthropologists made consistent measurements according to Gordon et al. (1989). Thirteen head and facial measurements were used in this study after considering important variables that contribute to racial distinctions, eliminating collinearities in the same craniofacial dimensions, and avoiding measurements with high inter-observer errors (e.g., head circumferences, stomion measurements). The majority of measurements, except for head length and breadth, were measured with an automated headboard device, calculated from three-dimensional ðX ; Y ; ZÞ coordinates of the corresponding landmarks. Head length and breadth were measured with a spreading caliper. The headboard units, tenths of millimeters, were converted into millimeters to be consistent with the caliper measurements.
Table 2 shows craniofacial measurements and their definitions, and Fig. 1 illustrates the anthropometric dimensions (Table 2) as demonstrated by Gordon et al. (1989). Canonical discriminant function analysis (SPSS, 1998, 8.0) was conducted to examine multivariate anthropometric differences between racial groups. The analysis selects the linear combinations of variables to maximize differences between groups by creating new axes called eigenvectors. Each individual was then assigned by discriminant scores in the multivariate anthropometric dimension. For comparing distributions of admixed groups against those in single-racial groups, discriminant scores of admixed groups were plotted against ellipses of Whites and Blacks, which were constructed from their discriminant scores (STATA, 1999, 7.0).
3. Results Table 3 gives the means and standard deviations (SD) of the craniofacial measurements by each racial category. Mean measurements of admixed groups such as mid-facial measurements (e.g., BIOBRMH, LIPLGTH,
Table 1 Distribution of age and race Race
Age o20 (28.9)a (28.2) (26.9) (37.5)
WM BM 1 2 MIXM 3 4 MIXM
237 335 7 3
Total
582 (28.5)
a
21–25
26–30
31–35
Total
303 400 13 3
178 272 4 0
102 182 2 2
820 1189 26 8
(37.0) (33.6) (50.0) (37.5)
719 (35.2)
(21.7) (22.9) (15.4) (0.0)
454 (22.2)
(12.4) (15.3) (7.7) (25.0)
288 (14.1)
(100.0) (100.0) (100.0) (100.0)
2043 (100.0)
Values in () indicate percentage values.
Table 2 Definitions of craniofacial measurements Abbreviation
Measurement
Definition
1. BIGONIAL 2. BIOCBRMH 3. BIZBDTH 4. CHINPROJ 5. HEADBRTH 6. HEADLGTH 7. LIPLGTH 8. MENSUBNH 9. NOSEBRTH 10. RTRAGX 11. SBNSSELH 12. SELLIONZ 13. SELTRAG
Bigonial breadth Biocular breadth Bizygomatic breadth Chin projection Head breadth Head length Lip length Menton–subnasal Nose breadth Right tragion X Subnasal-sellion length Sellion Z Facial projection
Straight-line between the right and left gonion on the jaw Distance between the right and left ectoorbitale Maximum horizontal breadth between zygomatic arches XYZ coordinates between right tragion and mentona Maximum horizontal breadth of the head Maximum distance between the glabella and back of the head Distance between the right and left cheilion on the corner of the mouth Distance between the menton and the subnasal Distance between the right and left alare Distance between right tragion and back of the head plane Distance between the subnasal and sellion Distance between sellion and top of the head plane Distance in XYZ coordinates between sellion and right tragiona
a
XYZ coordinates were not listed in Gordon et al. (1989).
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
381
MENSUBH, NOSEBRTH, SBNSSELH, SELLIONZ) fall between the White and Black groups, while the rest of the measurements did not follow the pattern. Facial breadths including lip length (LIPLGTH), nose breadth (NOSEBRTH), and biocular breadth (BIOCBRMH), as well as menton–subnasal height (MENSUBNH), show differences between Whites and Blacks. With relatively similar SD between racial groups, the result indicates that these mean measurements consistently tend to increase as Black heritage increases. On the other hand, when the SD among admixed groups is relatively similar to the SD for single racial groups, as is the case for mean nose height (SBNSSELH), those dimensions decrease as Black heritage increases. However, with large dispersions of head-related measurements (HEADBRTH HEADLGTH, RTRAGX, SELTRAG, SELLIONZ) and mandibular measurements (BIGONIAL, CHINPROJ, MENSUBNH), it is difficult to summarize how, in combination, craniofacial measurements reflect relationships among racial groups. The problem is particularly difficult, given the small sample population, for the admixed group with 34 Black heritage. Table 4 is a canonical discriminant summary including eigenvalues and the canonical correlation between discriminant scores and groups. Table 5 is the structure matrix indicating the weighting for each variable. Table 5 indicates the correlation between each measurement and its discriminant function. The results reveal a significant distinction between racial groups found only in the first eigenvalue ðpo0:05Þ (Table 4). The first eigenvalue shows 99.6% of the group variation with a high canonical correlation (0.84) (Table 4). The structure matrix indicates that nose breadth and lip length retain strong relationships to the first function. Plots and ellipses are constructed based on the first two discriminant scores. Function 1, the horizontal axis,
Fig. 1. The illustration of craniofacial measurements.
Table 3 Descriptive summary of craniofacial measurements (mm) Variables
BIGONIAL BIOCBRMH BIZBDTH CHINPROJ HEADBRTH HEADLGTH LIPLGTH MENSUBNH NOSEBRTH RTRAGX SBNSSELH SELLIONZ SELTRAG
White male ðN ¼ 820Þ
Black male ðN ¼ 1189Þ
Biracial male ðN ¼ 26Þ
3 4Black
Mean
SD
Mean
SD
Mean
SD
Mean
SD
117.6 120.3 139.3 149.6 150.9 197.3 54.4 72.4 33.9 100.0 51.1 112.6 124.5
7.3 5.2 5.3 6.0 5.1 6.7 3.7 5.4 2.7 6.5 3.6 7.0 4.7
118.4 124.3 140.8 152.3 151.1 197.6 59.0 76.9 42.3 102.4 48.5 110.8 122.6
7.8 5.2 5.4 6.0 5.4 6.4 3.8 5.2 3.2 6.2 3.6 7.0 4.6
116.6 121.2 137.7 151.4 148.6 196.0 56.3 75.9 38.5 100.1 48.8 110.9 123.1
4.8 5.0 5.5 6.0 5.8 6.6 3.1 4.4 3.2 5.9 2.9 6.7 5.8
117.7 122.2 137.5 149.5 148.4 194.5 57.7 76.2 41.3 98.8 48.6 112.3 120.3
10.1 4.0 4.5 8.3 5.6 10.8 2.7 5.6 2.7 8.2 2.6 9.3 7.8
male ðN ¼ 8Þ
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
382 Table 4 Canonical discriminant function summary Function
Eigenvalue
Percentage of variance
Canonical correlation
Chi-square (df)
Significant
1 2 3
2.47 0.01 0.00
99.6 0.3 0.1
0.84 0.10 0.06
2552.0 (39) 22.2 (24) 5.7 (11)
0.00 0.56 0.89
Table 5 Structure matrix
Variables
Function 1
Function 2
NOSEBRTH LIPLGTH MENSUBNH BIZBDTH HEADBRTH RTRAGX BIOCBRMH HEADLGTH SBNSSELH BIGONIAL CHINPROJ SELTRAG SELLIONZ
0.86 0.4 0.27 0.08 0.01 0.12 0.24 0.01 −0.22 0.03 0.14 −0.13 −0.08
0.13 0.29 −0.16 0.68 0.66 0.4 0.4 0.4 0.28 0.24 0.11 0.27 0.03
Note: Shaded area indicates strong correlation between each variable and discriminant function.
6
WM MlX12M
BM MlX34M
BM
function2
3 0
Large nose breadth and lip length -3 -6 -6 -3 0 3 function1 Note: MIX12M = Biracial male MIX34M = Individuals with 43 Black ancestry
6
Fig. 2. A two-dimensional plot for admixed males with ellipses of White (WM) and Black males (BM).
comprises the nose and lip breadth measurements, and these measurements increase as the rate of Black heritage increases. Function 2, the vertical axis, comprises the frame of craniofacial measurements, although it does not show strong anthropometric patterns. Fig. 2 is a twodimensional plot for admixed male groups, with ellipses for White and Black male populations. The distribution of male admixed groups fits within the single race distributions. The majority of biracial males are distributed in the area where the two ellipses overlap. The remainder of the individuals are located within either the Black or White ellipses, suggesting subjects’ self reported race may not always accurately reflect their biological
background. However, none of the individuals are located outside of single racial groups. The majority of admixed males with 34 Black ancestry are within the Black ellipse, and none is located in the White ellipse.
4. Discussion The background of the populations used to derive anthropometric data need to be carefully considered when designing and sizing equipment as anthropometric data are sensitive to the distribution of age, race, or occupation (Hsiao et al., 2002). In this study, a wide range of subject ages was not studied, thus the chronological changes in craniofacial anthropometry within subjects were relatively small. This study examined the anthropometric distributions of admixed groups and single racial groups, and their affects on ergonomic designs. The descriptive results showed distinctions between racial groups in mid-facial measurements, yet they did not show distinction in how, in combination, craniofacial measurements reflected relationships among racial groups. Multivariate results confirmed that mid-facial measurements such as nose breadth and lip length, do differ between White and Black males. The anthropometric differences between Black and White Americans in the nasal area, as compared to overall facial areas, were also found in a previous study (Porter, 2004). Despite the small magnitudes in differences, mid-facial measurements including lip length and nose breadth may be considered when designing improved respirators, at least for these racial groups. In addition, problems were encountered during test fitting of Koreans with respirators designed primarily from White male’s anthropometry (Kim et al., 2003). The results of this study can be considered for head and face ergonomic designs for the US Army. Larger openings in respirators and chemical–biological protective masks than would be necessary for Whiteonly populations may allow better air flow through the nose and mouth of Black individuals. Anthropometric investigations, along with fit tests, will also enhance occupational performance and endurance, assist in the prevention from worker’s discomfort, and reduce potential health hazards. The multivariate approach summarized the distribution of anthropometric measurements between racial
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
groups, based on discriminant scores calculated from the linear combinations of an individual’s measurements, as a multivariate dimension. Among admixed groups, their multivariate anthropometric distributions are within single-race distributions. The results of this study indicate that additional sizing or adjustments for admixed populations are not required if anthropometric distributions of single-race populations that comprise admixed populations are known. Thus changes in admixed population numbers can be accommodated by changing the distribution or tariff of clothing and mask sizes in the logistics system rather than adjusting the clothing designs and adding new sizing to accommodate new or unique anthropometric combinations. Due to the low number of admixed individuals in our pool of prospective volunteers, the sample size was very small in this preliminary study. A larger sample size will be required to improve and validate the results in a future study. The number of biologically admixed populations is increasing in the US (Gibson and Lennon, 1999). This type of ergonomic study can be used to assess future designs and sizing to accommodate the growing admixed populations. For ergonomic concerns, this preliminary study showed that the craniofacial anthropometry of biologically admixed groups (in this case, of Blacks and Whites) falls within the range defined by single racial populations. The approach shown in this study is not limited to the head and face or only military equipment. The same approach may be applied to other occupations or industries (e.g., fire fighters, workers in chemical industry), designs that require anthropometric standards, and other racial/ethnic groups. Finally, any gender effect is unknown. Thus, the question of whether or not the results in this study are applicable to a population of admixed females should also be examined. Acknowledgements The author would like to thank Dr. Claire Gordon, US Army Soldier Systems Center for providing the
383
ANSUR data sets. My appreciation also goes to Mr. John Roebuck, Roebuck Research and Consulting, Dr. William Santee, US Army Research Institute of Environmental Medicine (USARIEM), and Ms. Shari Hallas, USARIEM, for reviewing this paper. The investigators have adhered to the policies for protection of human subjects as prescribed in Army Regulation 7025, and the research was conducted in adherence with the provision of 45 CFR Part 46. The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Army or the Department of Defense.
References Bradtmiller, B., Ratnaparkhi, J., Tebbetts, I., 1985. Demographic and anthropometric assessment of US Army anthropometric database. NATICK/TR-86/004. US Army Natick Research, Development, and Engineering Center, Natick, MA. Brues, A., 1945. Study of Anthropometric Data. Massachusetts Institute of Technology, Cambridge, MA. Gibson, C.J., Lennon, E., 1999. Historical census statistics on the foreign-born population of the United States: 1850–1990. Population division. US Bureau of the Census, Washington, DC. Gordon, C.C., 1996. Downsizing, demographic change, and validity of the 1988 ANSUR data in 1966. NATICK/TR-97/003. US Army Natick Research, Development, and Engineering Center, Natick, MA. Gordon, C.C., Churchill, T., Clauser, C.E., Bradtmiller, B., McConville, J.T., Tebbetts, I., Walker, R.A., 1989. 1988 Anthropometric survey of US Army personnel: methods and summary statistics. NATICK/TR-89/044. US Army Natick Research, Development, and Engineering Center, Natick, MA. Hsiao, H., Long, D., Snyder, K., 2002. Anthropometric differences among occupational groups. Ergonomics 45, 136–152. Kim, H., Han, D., Roh, Y., Kim, K., Park, Y., 2003. Facial anthropometric dimensions of Koreans and their associations with fit of quarter-mask respirators. Ind. Health 41, 8–18. Porter, J., 2004. The average African American male face. Arch. Facial Plast. Surg. 6, 78–81. SPSS, 1998. SPSS Base 8.0. Chicago: SPSS. STATA, 1999. STATA Release 7.0 College Station, Stata Cooperation.
Applied Ergonomics 36 (2005) 379–383 www.elsevier.com/locate/apergo
Technical note
Head and facial anthropometry of mixed-race US Army male soldiers for military design and sizing: A pilot study Miyo Yokota Geo-Centers, Inc., 190 N. Main Street, Natick, MA 01760, USA Received 5 January 2004; accepted 14 January 2005
Abstract In the United States, the biologically admixed population is increasing. Such demographic changes may affect the distribution of anthropometric characteristics, which are incorporated into the design of equipment and clothing for the US Army and other large organizations. The purpose of this study was to examine multivariate craniofacial anthropometric distributions between biologically admixed male populations and single racial groups of Black and White males. Multivariate statistical results suggested that nose breadth and lip length were different between Blacks and Whites. Such differences may be considered for adjustments to respirators and chemical–biological protective masks. However, based on this pilot study, multivariate anthropometric distributions of admixed individuals were within the distributions of single racial groups. Based on the sample reported, sizing and designing for the admixed groups are not necessary if anthropometric distributions of single racial groups comprising admixed groups are known. r 2005 Elsevier Ltd. All rights reserved. Keywords: Head and face anthropometry
1. Introduction Anthropometry has been critical to successfully designing and sizing protective helmets, eyewear, and respirators in the US Army. Compared to early demography based primarily on White males, the anthropometric distribution of the US Army has changed since individuals of diverse ethnic backgrounds began serving in the 1960s (Bradtmiller et al., 1985; Brues, 1945; Gordon, 1996). Such demographic changes may affect the anthropometric distributions of the populations for which military equipment was designed. Current US Army ergonomics are based on the 1988 US Army anthropometric survey (ANSUR) database (Gordon et al., 1989), weighted for primary racial groups and different age proportions in each gender throughout US Army demography. However, despite the fact that the Tel.: +1 508 233 5845; fax: +1 508 233 5298.
E-mail addresses: [email protected], [email protected]. 0003-6870/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.apergo.2005.01.009
number of biologically admixed populations is increasing in the US (Gibson and Lennon, 1999), there have been few related ergonomic studies. This study investigated differences in multivariate head and face anthropometric distributions between biologically admixed male individuals and single racial male groups (Whites or Blacks). The goal of this study was to examine the effect that an increasingly admixed US Army population might have on current and future equipment sizing and design statistics.
2. Materials and methods Craniofacial anthropometry of biracial male soldiers ðN ¼ 26Þ and admixed soldiers with 14 White ancestry and 34 Black ðN ¼ 8Þ were compared to those of White soldiers ðN ¼ 820Þ and Black soldiers ðN ¼ 1189Þ: Race was determined during interviews with an anthropologist by a subject’s report of ancestry and his parents’ ancestries. The ANSUR survey contains separate
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
380
questions on race and ethnicity because people are more familiar with the term ‘‘race’’ to describe their biological background. All subjects were born in the US. Their ages ranged between 18 and 35 years, and were grouped into 5-year intervals (below 20, 21–25, 26–30, 31–35). The age distribution between White, Black, and mixed groups was homogenous ðp40:05Þ; based on a Fisher’s exact test. Table 1 shows the summary of age distribution by racial groups. Anthropologists made consistent measurements according to Gordon et al. (1989). Thirteen head and facial measurements were used in this study after considering important variables that contribute to racial distinctions, eliminating collinearities in the same craniofacial dimensions, and avoiding measurements with high inter-observer errors (e.g., head circumferences, stomion measurements). The majority of measurements, except for head length and breadth, were measured with an automated headboard device, calculated from three-dimensional ðX ; Y ; ZÞ coordinates of the corresponding landmarks. Head length and breadth were measured with a spreading caliper. The headboard units, tenths of millimeters, were converted into millimeters to be consistent with the caliper measurements.
Table 2 shows craniofacial measurements and their definitions, and Fig. 1 illustrates the anthropometric dimensions (Table 2) as demonstrated by Gordon et al. (1989). Canonical discriminant function analysis (SPSS, 1998, 8.0) was conducted to examine multivariate anthropometric differences between racial groups. The analysis selects the linear combinations of variables to maximize differences between groups by creating new axes called eigenvectors. Each individual was then assigned by discriminant scores in the multivariate anthropometric dimension. For comparing distributions of admixed groups against those in single-racial groups, discriminant scores of admixed groups were plotted against ellipses of Whites and Blacks, which were constructed from their discriminant scores (STATA, 1999, 7.0).
3. Results Table 3 gives the means and standard deviations (SD) of the craniofacial measurements by each racial category. Mean measurements of admixed groups such as mid-facial measurements (e.g., BIOBRMH, LIPLGTH,
Table 1 Distribution of age and race Race
Age o20 (28.9)a (28.2) (26.9) (37.5)
WM BM 1 2 MIXM 3 4 MIXM
237 335 7 3
Total
582 (28.5)
a
21–25
26–30
31–35
Total
303 400 13 3
178 272 4 0
102 182 2 2
820 1189 26 8
(37.0) (33.6) (50.0) (37.5)
719 (35.2)
(21.7) (22.9) (15.4) (0.0)
454 (22.2)
(12.4) (15.3) (7.7) (25.0)
288 (14.1)
(100.0) (100.0) (100.0) (100.0)
2043 (100.0)
Values in () indicate percentage values.
Table 2 Definitions of craniofacial measurements Abbreviation
Measurement
Definition
1. BIGONIAL 2. BIOCBRMH 3. BIZBDTH 4. CHINPROJ 5. HEADBRTH 6. HEADLGTH 7. LIPLGTH 8. MENSUBNH 9. NOSEBRTH 10. RTRAGX 11. SBNSSELH 12. SELLIONZ 13. SELTRAG
Bigonial breadth Biocular breadth Bizygomatic breadth Chin projection Head breadth Head length Lip length Menton–subnasal Nose breadth Right tragion X Subnasal-sellion length Sellion Z Facial projection
Straight-line between the right and left gonion on the jaw Distance between the right and left ectoorbitale Maximum horizontal breadth between zygomatic arches XYZ coordinates between right tragion and mentona Maximum horizontal breadth of the head Maximum distance between the glabella and back of the head Distance between the right and left cheilion on the corner of the mouth Distance between the menton and the subnasal Distance between the right and left alare Distance between right tragion and back of the head plane Distance between the subnasal and sellion Distance between sellion and top of the head plane Distance in XYZ coordinates between sellion and right tragiona
a
XYZ coordinates were not listed in Gordon et al. (1989).
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
381
MENSUBH, NOSEBRTH, SBNSSELH, SELLIONZ) fall between the White and Black groups, while the rest of the measurements did not follow the pattern. Facial breadths including lip length (LIPLGTH), nose breadth (NOSEBRTH), and biocular breadth (BIOCBRMH), as well as menton–subnasal height (MENSUBNH), show differences between Whites and Blacks. With relatively similar SD between racial groups, the result indicates that these mean measurements consistently tend to increase as Black heritage increases. On the other hand, when the SD among admixed groups is relatively similar to the SD for single racial groups, as is the case for mean nose height (SBNSSELH), those dimensions decrease as Black heritage increases. However, with large dispersions of head-related measurements (HEADBRTH HEADLGTH, RTRAGX, SELTRAG, SELLIONZ) and mandibular measurements (BIGONIAL, CHINPROJ, MENSUBNH), it is difficult to summarize how, in combination, craniofacial measurements reflect relationships among racial groups. The problem is particularly difficult, given the small sample population, for the admixed group with 34 Black heritage. Table 4 is a canonical discriminant summary including eigenvalues and the canonical correlation between discriminant scores and groups. Table 5 is the structure matrix indicating the weighting for each variable. Table 5 indicates the correlation between each measurement and its discriminant function. The results reveal a significant distinction between racial groups found only in the first eigenvalue ðpo0:05Þ (Table 4). The first eigenvalue shows 99.6% of the group variation with a high canonical correlation (0.84) (Table 4). The structure matrix indicates that nose breadth and lip length retain strong relationships to the first function. Plots and ellipses are constructed based on the first two discriminant scores. Function 1, the horizontal axis,
Fig. 1. The illustration of craniofacial measurements.
Table 3 Descriptive summary of craniofacial measurements (mm) Variables
BIGONIAL BIOCBRMH BIZBDTH CHINPROJ HEADBRTH HEADLGTH LIPLGTH MENSUBNH NOSEBRTH RTRAGX SBNSSELH SELLIONZ SELTRAG
White male ðN ¼ 820Þ
Black male ðN ¼ 1189Þ
Biracial male ðN ¼ 26Þ
3 4Black
Mean
SD
Mean
SD
Mean
SD
Mean
SD
117.6 120.3 139.3 149.6 150.9 197.3 54.4 72.4 33.9 100.0 51.1 112.6 124.5
7.3 5.2 5.3 6.0 5.1 6.7 3.7 5.4 2.7 6.5 3.6 7.0 4.7
118.4 124.3 140.8 152.3 151.1 197.6 59.0 76.9 42.3 102.4 48.5 110.8 122.6
7.8 5.2 5.4 6.0 5.4 6.4 3.8 5.2 3.2 6.2 3.6 7.0 4.6
116.6 121.2 137.7 151.4 148.6 196.0 56.3 75.9 38.5 100.1 48.8 110.9 123.1
4.8 5.0 5.5 6.0 5.8 6.6 3.1 4.4 3.2 5.9 2.9 6.7 5.8
117.7 122.2 137.5 149.5 148.4 194.5 57.7 76.2 41.3 98.8 48.6 112.3 120.3
10.1 4.0 4.5 8.3 5.6 10.8 2.7 5.6 2.7 8.2 2.6 9.3 7.8
male ðN ¼ 8Þ
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
382 Table 4 Canonical discriminant function summary Function
Eigenvalue
Percentage of variance
Canonical correlation
Chi-square (df)
Significant
1 2 3
2.47 0.01 0.00
99.6 0.3 0.1
0.84 0.10 0.06
2552.0 (39) 22.2 (24) 5.7 (11)
0.00 0.56 0.89
Table 5 Structure matrix
Variables
Function 1
Function 2
NOSEBRTH LIPLGTH MENSUBNH BIZBDTH HEADBRTH RTRAGX BIOCBRMH HEADLGTH SBNSSELH BIGONIAL CHINPROJ SELTRAG SELLIONZ
0.86 0.4 0.27 0.08 0.01 0.12 0.24 0.01 −0.22 0.03 0.14 −0.13 −0.08
0.13 0.29 −0.16 0.68 0.66 0.4 0.4 0.4 0.28 0.24 0.11 0.27 0.03
Note: Shaded area indicates strong correlation between each variable and discriminant function.
6
WM MlX12M
BM MlX34M
BM
function2
3 0
Large nose breadth and lip length -3 -6 -6 -3 0 3 function1 Note: MIX12M = Biracial male MIX34M = Individuals with 43 Black ancestry
6
Fig. 2. A two-dimensional plot for admixed males with ellipses of White (WM) and Black males (BM).
comprises the nose and lip breadth measurements, and these measurements increase as the rate of Black heritage increases. Function 2, the vertical axis, comprises the frame of craniofacial measurements, although it does not show strong anthropometric patterns. Fig. 2 is a twodimensional plot for admixed male groups, with ellipses for White and Black male populations. The distribution of male admixed groups fits within the single race distributions. The majority of biracial males are distributed in the area where the two ellipses overlap. The remainder of the individuals are located within either the Black or White ellipses, suggesting subjects’ self reported race may not always accurately reflect their biological
background. However, none of the individuals are located outside of single racial groups. The majority of admixed males with 34 Black ancestry are within the Black ellipse, and none is located in the White ellipse.
4. Discussion The background of the populations used to derive anthropometric data need to be carefully considered when designing and sizing equipment as anthropometric data are sensitive to the distribution of age, race, or occupation (Hsiao et al., 2002). In this study, a wide range of subject ages was not studied, thus the chronological changes in craniofacial anthropometry within subjects were relatively small. This study examined the anthropometric distributions of admixed groups and single racial groups, and their affects on ergonomic designs. The descriptive results showed distinctions between racial groups in mid-facial measurements, yet they did not show distinction in how, in combination, craniofacial measurements reflected relationships among racial groups. Multivariate results confirmed that mid-facial measurements such as nose breadth and lip length, do differ between White and Black males. The anthropometric differences between Black and White Americans in the nasal area, as compared to overall facial areas, were also found in a previous study (Porter, 2004). Despite the small magnitudes in differences, mid-facial measurements including lip length and nose breadth may be considered when designing improved respirators, at least for these racial groups. In addition, problems were encountered during test fitting of Koreans with respirators designed primarily from White male’s anthropometry (Kim et al., 2003). The results of this study can be considered for head and face ergonomic designs for the US Army. Larger openings in respirators and chemical–biological protective masks than would be necessary for Whiteonly populations may allow better air flow through the nose and mouth of Black individuals. Anthropometric investigations, along with fit tests, will also enhance occupational performance and endurance, assist in the prevention from worker’s discomfort, and reduce potential health hazards. The multivariate approach summarized the distribution of anthropometric measurements between racial
ARTICLE IN PRESS M. Yokota / Applied Ergonomics 36 (2005) 379–383
groups, based on discriminant scores calculated from the linear combinations of an individual’s measurements, as a multivariate dimension. Among admixed groups, their multivariate anthropometric distributions are within single-race distributions. The results of this study indicate that additional sizing or adjustments for admixed populations are not required if anthropometric distributions of single-race populations that comprise admixed populations are known. Thus changes in admixed population numbers can be accommodated by changing the distribution or tariff of clothing and mask sizes in the logistics system rather than adjusting the clothing designs and adding new sizing to accommodate new or unique anthropometric combinations. Due to the low number of admixed individuals in our pool of prospective volunteers, the sample size was very small in this preliminary study. A larger sample size will be required to improve and validate the results in a future study. The number of biologically admixed populations is increasing in the US (Gibson and Lennon, 1999). This type of ergonomic study can be used to assess future designs and sizing to accommodate the growing admixed populations. For ergonomic concerns, this preliminary study showed that the craniofacial anthropometry of biologically admixed groups (in this case, of Blacks and Whites) falls within the range defined by single racial populations. The approach shown in this study is not limited to the head and face or only military equipment. The same approach may be applied to other occupations or industries (e.g., fire fighters, workers in chemical industry), designs that require anthropometric standards, and other racial/ethnic groups. Finally, any gender effect is unknown. Thus, the question of whether or not the results in this study are applicable to a population of admixed females should also be examined. Acknowledgements The author would like to thank Dr. Claire Gordon, US Army Soldier Systems Center for providing the
383
ANSUR data sets. My appreciation also goes to Mr. John Roebuck, Roebuck Research and Consulting, Dr. William Santee, US Army Research Institute of Environmental Medicine (USARIEM), and Ms. Shari Hallas, USARIEM, for reviewing this paper. The investigators have adhered to the policies for protection of human subjects as prescribed in Army Regulation 7025, and the research was conducted in adherence with the provision of 45 CFR Part 46. The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Army or the Department of Defense.
References Bradtmiller, B., Ratnaparkhi, J., Tebbetts, I., 1985. Demographic and anthropometric assessment of US Army anthropometric database. NATICK/TR-86/004. US Army Natick Research, Development, and Engineering Center, Natick, MA. Brues, A., 1945. Study of Anthropometric Data. Massachusetts Institute of Technology, Cambridge, MA. Gibson, C.J., Lennon, E., 1999. Historical census statistics on the foreign-born population of the United States: 1850–1990. Population division. US Bureau of the Census, Washington, DC. Gordon, C.C., 1996. Downsizing, demographic change, and validity of the 1988 ANSUR data in 1966. NATICK/TR-97/003. US Army Natick Research, Development, and Engineering Center, Natick, MA. Gordon, C.C., Churchill, T., Clauser, C.E., Bradtmiller, B., McConville, J.T., Tebbetts, I., Walker, R.A., 1989. 1988 Anthropometric survey of US Army personnel: methods and summary statistics. NATICK/TR-89/044. US Army Natick Research, Development, and Engineering Center, Natick, MA. Hsiao, H., Long, D., Snyder, K., 2002. Anthropometric differences among occupational groups. Ergonomics 45, 136–152. Kim, H., Han, D., Roh, Y., Kim, K., Park, Y., 2003. Facial anthropometric dimensions of Koreans and their associations with fit of quarter-mask respirators. Ind. Health 41, 8–18. Porter, J., 2004. The average African American male face. Arch. Facial Plast. Surg. 6, 78–81. SPSS, 1998. SPSS Base 8.0. Chicago: SPSS. STATA, 1999. STATA Release 7.0 College Station, Stata Cooperation.