Infant, Child and Teenager Anthropometry For Product Safety Design

As of 1972 a source of reliable and consistent data providing measurement statistics for both traditional and functional anthropometric measures for a sample representative of the infants, children, and youth in the United States was not available. This was surprising in view of the fact that some 800 studies in the literature were found to provide child measurement data (Snyder et al, 1973; reprinted 1975). This comprehensive review of the literature on child and infant anthropometric data revealed that while a large number of studies had been conducted, the type of measurements needed for most safety applications were actually non-existent.

A large number of these studies were longitudinal (rather than cross-sectional) surveys designed to study the growth of a fixed population over a period of time using only a few basic measures such as height and weight. All the studies were limited in that they either used specific populations, selected age ranges, or only a few measurements, and did not consider differences due to region of the country, race, and other environmental and socio-economic factors likely to affect the sizes of children. Even if these factors had been considered and if all the measurements necessary had been taken in one study or another, differences due to measurer technique, definitions in measurements, and differences in age groupings would make it impossible to compile these data into one common source of reliable and consistent information. Additionally, many growth studies as recently as Marshall and Carter (1975) have found that children today are taller and larger than children their same age in past generations. Since two-thirds of the 35 most complete studies encountered had been made prior to 1960, it is questionable how representative much of these data may be of the current population of U.S. children.

This point was emphasized in a 1975 study conducted jointly by physical anthropologists from several organizations attempting to evaluate masterbody forms for 3-year and 6-year old child dummies for the Federal Aviation Administration (Young et al, 1975) and National Highway Traffic Safety Administration (Reynolds et al, 1975). It was found that of 98 body dimensions determined to be necessary for describing the external morphology of a child test manikin, 68 were not available in the literature. For example, the shape of the torso at the nipples is described by circumference, breadth, depth, and height from seat pan, but only circumference and depth measurements had been previously taken. Unfortunately this work was conducted before data were available from the initial CPSC Study. Problems related to anatomical considerations for the design of child restraints, another application where anthropometric data were needed, has been outlined by Burdi et al (1969).

OBJECTIVES

The purpose of the two studies conducted for CPSC was to provide anthropometric data of both a functional and traditional nature to describe the sizes, shapes, body linkage, selected functional measures, and seated and standing centers of gravity for a population of infants, children, and teenagers representative of the 48 conterminous United States. Tasks involved determining the measurements most required; the experimental design; designing, constructing, and testing unique advanced anthropometric instrumentation including mini-computer systems and entirely new center of gravity devices; collection, reduction, and analysis of the required dimensional data; and finally, tabulation and representation of the data in a format most useful to persons concerned with designing or establishing guidelines and standards for children's products, taking into consideration the potential functional applications.

METHODS

This study was undertaken as a multi-disciplinary team effort with co-investigators representing specialties in physical anthropology, bioengineering, pediatrics, computer engineering, industrial and operations engineering and biostatistics. In addition, advisory consultation with other scientists occurred as the study progressed.

To establish the numbers of various types of infants and children to be included in the sample, the initial study utilized U.S. Census and HEW data to project a sample of 4000 subjects whose racial and socio-economic characteristics matched the proportions of those characteristics in the total U.S. population. While the 1972-1975 study involved 41 measurements on infants and children from age 2 weeks to 13 years, the 1975-1977 study expanded this data base to also include 13 to 18 year olds (a population for which little data were known), and expanded the number of measurements to 87.

To assure the national representativeness of school-age subjects, a sample of schools was selected from a national list available from the third round of a study of youth fitness completed in the Department of Physical Education at the University of Michigan. The selection process used in the youth fitness study has been described in detail in Reiff, et al. (1968). The U-M Survey Research Center maintains demographic and socioeconomic data on a national sample of geographic-political units representative of the U.S. population. Utilizing this sampling frame as a base, a smaller sampling frame was conducted for the anthropometric study. It consisted of random selection of eight central-city areas, 10 suburban areas, and 31 areas outside large metropolitan areas. The next step involved random selection of one school district within each of the sampled areas. Finally, specific schools in those districts were randomly selected, using a ratio of two elementary schools to one intermediate school to one senior high school. Because of scheduling and space problems, some schools originally selected did not participate in the measurement study. However, every effort was made to select matching schools from the same area.

Measurements and Measurement Strategy

Traditionally, anthropometric measurements have been devised to describe human variation in body morphology, body proportions, and changes in size attributed to growth, race, and other variables of interest to anthropologists. These dimensions were taken to describe the linear distance between two landmarks (heights, breadths, and depths) or around a body segment at a prescribed level (circumference). In today's world, however, this classical approach to anthropometry is not sufficient to answer questions relevant to the interaction of man with a multitude of man-made environments. While the traditional anthropometry measures are still important as they describe differences between individuals and populations, the differences must also be considered as a part of a man-machine system. Thus a systems approach to anthropometry was used as a general guiding principle. Consideration was also given to the probability that another study of this magnitude would not be likely in the near future and it was thus essential to provide the maximum amount of information possible within the budget and time limitations and to identify those measurements that would be of most interest and usefulness to the greatest number of potential users.

The result of these considerations was a list of 87 measurements shown on the following page. Traditionally, all measurements are made on every subject in the survey sample. However, because time available per subject was limited to 15-20 minutes so as to assure school and subject cooperation, this permitted obtaining only about 40 measurements per subject. The strategy adopted was to obtain the 22 Core measurements (Group I) on all subjects, plus Group II, Group III, or Group IV measurements sequentially, on every third subject. Thus, 42-45 measurements were taken on each subject, and the sample sizes for measurements in Groups II, III, and IV are one-third as large as for Group I.

The key factor in this approach was to select measurements for Group I that correlate highly with measurements in the other three groups. Because the core measurements were taken on every subject, and each of the non-core measurements correlates highly with one or more of the core measurements, this provided a means of checking the representativeness and comparability of the non-core sample populations. Statistical comparisons of measurement means in selected core measurements and age groups showed that there is no significant difference between the samples for Groups II, III, and IV.

To avoid any systematic association of measurements with subjects in a particular age group or region of the U.S. all measurement sets were taken equally at all measurement sites. To accomplish this, the combination of measurement sets (I + II, I + III, I + IV) were done sequentially.

Measurements, Age 2-19

I. Core Measurements. Weight, Stature, Head Circumference, Chest Circumference, Waist Circumference, Hip Circumference, Upper Arm Circumference, Forearm Circumference, Upper Thigh Circumference, Calf Circumference, Foot length, Foot Breadth, Hand Length, Head Breadth, Shoulder Breadth, Shoulder-Elbow Length, Elbow-Hand Length, Maximum Hip Breadth (sitting), Buttock-Knee Length, Knee Height, Erect Sitting Height, Head Breadth.

II. Shape Measurements. Chest Height, Waist Height, Hip Height, Chest Breadth, Waist Breadth, Neck Circumference, Natural Waist Circumference, Wrist Circumference, Ankle Circumference, Calf Circumference, Ankle Breadth, Calf Breadth, Upper Thigh Depth, Wrist Breadth, Forearm Depth, Upper Arm Depth, Neck Breadth, Head Length, Eye Height, Thigh Clearance, Maximum Thigh Breadth (sitting).

III. Linkage Measurements. Vertical Arm Reach, Frontal Arm Reach, Lateral Arm Reach, Step Height, Sup-rasternale Height, Iliocristale Height, Iliospinale Height, Trochanteric Height, Gluteal Furrow Height, Sphyrion Height, Tibiale Height, Bispinous Breadth, Clavicale-Acromion Length, Acromion-Radiale Length, Ra-diale-Stylion Length, Biacromial Breadth, Hip Breadth at Trochanter, Supine Stature, Supine Sitting Height, Standing Center of Gravity, Seated Center of Gravity.

IV. Head, Face, and Hand Measurements. Bizygoma-tic Breadth, Maximum Frontal Breadth, Mouth Breadth, Nose Length, Lower Face Height, Face Height, Head Height, Tragion to Back of Head, Tragion to Top of Head, Ear-Sellion Depth, Bitragion Breadth, Ball of Fist Circumference, Maximum Fist Breadth, Maximum Fist Depth, Thumb-Crotch-Middle Finger Length, Middle Finger-Thumb Grip Length, Thumb Length, Index Finger Length, Middle Finger Length, Middle Finger Diameter, Index Finger Diameter, Thumb Diameter, Minimum Hand Clearance.

Infant Measurements. Since infants (age range two weeks to 24 months) do not have mature skeletal and muscular systems and cannot perform movements required for most of the functional measurements, a separate set of 34 measurements was compiled and used for these subjects. These were:

Weight, Crown-Sole Length, Crown-Rump Length, Head Circumference, Head Breadth, Head Length, Shoulder Breadth, Shoulder-Elbow Length, Upper Arm Circumference, Elbow-Hand Length, Forearm Circumference, Wrist Circumference, Hand Length, Hand Breadth, Minimum Hand Clearance, Maximum Fist Breadth, Thumb Diameter, Middle Finger Diameter, Chest Circumference, Chest Breadth, Waist Circumference, Waist Breadth, Rump-Sole Length, Rump-Knee Length, Hip Circumference, Hip Breadth, Mid-Thigh Circumference, Mid-Thigh Depth, Knee-Sole Length, Calf Circumference, Ankle Circumference, Ankle Breadth, Foot Length, Foot Breadth.

DATA COLLECTION

In the 1972-1975 study, measurements were made at 71 schools, clinics, and day care centers in 8 states. The 1975-1977 study included measurements obtained at 104 locations in eight regions (20 states) by two teams. In the latter study, one consisted of two males, the other of two females. Team members were selected for their ability and experience in working with children and teenagers. They received extensive training and supervised practice before commencing the field work. Several members were fluent in Spanish which was important in some schools particularly in the Southwest where English was a second language. In some cases materials to parents were printed in both Spanish and English. The process of obtaining approval from school authorities and parents, and of scheduling specific grades and classes, usually required considerable preliminary contacts. Ordinarily the school superintendent was approached by letter and then telephone. The request was presented to the school board for approval and then principals of schools in that district were contacted. Often the team coordinator visited the schools to explain the study to teachers and students, using an 18-minute film (University of Michigan, 1975) produced during the previous CPSC-sponsored study. After it was decided that a school would participate, an explanatory letter, consent form, and demographic questionnaire were distributed to parents of children in school grades selected to participate. The entire process of obtaining approval, informing teachers, parents, and students, and obtaining signed consent forms usually required about two months. When schools were closed during summer vacations, subjects were measured at summer camps.

The primary sources for subjects two to five years old were nursery schools and day care centers. Nationwide, infants were the most difficult age group to obtain, and additional sources for infant subjects were The University of Michigan Well Baby Clinic and a pediatric clinic in Pontiac, Michigan.

INSTRUMENTATION

An important aspect of this study was the design, fabrication and use of a portable mini-computer system for automatic retrieval of measurement data. Figure 1 illustrates this system and shows the NOVA 1220 computer system including 24 K of memory, 2 LINC type drive units 16 channels of A/D input, a signal conditioning and power supply package for processing instrument signals and activating device transducers, and an interactive display system consisting of keyboard, terminal and video display.

FIGURE 1

THE AUTOMATED ANTHROPOMETRIC SYSTEM

Four different automated measuring devices were interfaced with the NOVA computer system and utilized in the measuring process. These have been reported in detail by Snyder, et al. (1972; 1975), and Owings et al. (1974 a; 1974 b; 1974 c).

While computerized techniques have been suggested and explored previously (Garn, 1962; Prahl-Anderson, 1972; Bullock, 1974 [Margaret I. Bullock, University of Queensland, Brisbane, Australia (personal communication, 1974)]), the automated anthropometric system described is believed to be the most extensive systematic use of automated anthropometry, the first such use of the NOVA mini-computer data acquisition system, and the first practical portable computerized means of obtaining center of gravity measures on both infants and children.

For measuring linear distances (i.e., breadths, lengths, depths, heights) the automated caliper or anthropometer was used. These instruments are illustrated in the drawings of Figure 2. GPM (Siber Hegner & Co. Ltd) anthropometers and calipers were modified to provide electrical readout of length by means of a 10-turn potentiometer connected to the moving blade by a pulley and cable system. A miniature pressure transducer (Konigsberg Model P21) mounted in the special Plexiglas blade provided a means of standardizing measurements on soft tissue. Body linkage and head, face, and hand measurements were taken using removable pointed extensions of different lengths. A button in the stationary Plexiglas blade activates the measuring process on each device.

FIGURE 2

MODIFIED SLIDING CALIPERS AND MODIFIED ANTHROPOMETERS SHOWING POTENTIOMETER, PULLEY AND CABLE SYSTEM, AND PRESSURE TRANSDUCER.

A third automated device shown in Figure 3 was used for measuring various body circumferences. This consists of a standard steel measuring tape wrapped around a pulley which is attached to the shaft of a 10-turn potentiometer. The tape is wrapped about the particular body segment being measured and clipped back on the tip of the aluminum tape guide. Tension is provided by a coil spring located inside the pulley and the measurement is sensed by the computer upon depressing the plastic button with the thumb or index finger.

FIGURE 3

MODIFIED GIRTH DEVICE FOR ELECTRICAL READOUT OF CIRCUMFERENCE MEASUREMENTS

Two devices were designed for measuring whole body center of gravity, one for infants and one for children and teenagers up to 250 pounds. The principal of operation is the same for both and is illustrated for the larger device in Figure 4. The subject is placed in either the supine standing or sitting position on a rigid platform supported by three precisely calibrated load cells. Outputs of the load transducing cells are amplified before entering the A/D unit of the computer. By sensing the load cell readings with the subject in position and with the device empty, the computer calculates the location of the center of gravity relative to a known reference plane.

FIGURE 4

CENTER OF GRAVITY DEVICE WITH SUBJECT IN SUPINE STANDING (UPPER) AND SUPINE SITTING POSITION (LOWER)

Two complete systems including computer, measurement instruments, TV display, and center of gravity devices were assembled and transported throughout the United States by the two teams in Dodge Maxi-vans as illustrated in Figure 5.

Other devices used for taking other functional measurements include a hole template for measuring finger and thumb diameters, hinged Plexiglas hole boards for measuring hand clearance (Figure 6) and a grip rod with pivoting extension for measuring functional grip distances (not shown). In order to standardize and facilitate the positioning of the subjects for various measurements, a portable table and adjustable foot rest platform were designed and used with each measuring system (See Figure 1).

FIGURE 5

MAXI-VAN USED TO TRANSPORT ANTHROPOMETRIC EQUIPMENT

FIGURE 6

HOLE TEMPLATE FOR MEASURING FINGER DIAMETER AND HINGED PLEXIGLAS HOLE BOARDS FOR MEASURING HAND CLEARANCE.

The entire measuring process including sequencing through the measurement data sets, calibration of automated instruments and pressure transducers, recording of subject demographic information, retrieving and recording measurement values from the various automated devices, checking measurement values relative to subject age and stature, and calculating subject center of gravity was controlled by a computer program called "MAP" or Michigan Anthropometric Processor.

DATA HANDLING, EDITING, AND ANALYSIS

At the completion of each geographic location data tapes were returned to HSRI for transfer to a file on the Michigan Terminal System (MTS). Interim editing took place during the course of the study by listing measurement values according to subject age and visually checking for bad data. Final editing was accomplished when data collection was complete by using two routines written for the HSRI PDP 11/45 system.

A program called X-VAL, obtained from anthropologists at Wright Patterson AFB, was used to search for extreme measurement values in designated age groups. Subjects were grouped by data set, sex, and age intervals of one year each, and the program listed the ten smallest and ten largest values for each measurement along with statistics of the distribution (e.g., mean, standard deviation, skewness). Visual inspection of the data was then used to search for subject numbers with unreasonable measurement values and a complete listing of the data on these subjects was produced. From this listing a decision could be made as to whether to delete or leave the value in question.

A second procedure used to edit the data was multiple regression analysis. This involved establishing a linear relationship between one measurement variable and several other measurement variables. For example, one might attempt to establish a linear relationship of weight with stature, chest circumference, and waist circumference [weight = f (stature, chest circ., waist circ.)]. Once a relationship was established, measurement values which fell outside the predicted value, plus or minus specified error tolerances, were edited appropriately. This procedure allowed for editing points which were incorrect but which still fell within two standard deviations of the mean measurement value for a given age range (i.e., they are not extreme values).

SAMPLE POPULATION CHARACTERISTICS

The composition of the sample population of measurement subjects by race and sex is shown in Table 1. As the table shows, 50.7 percent of the subjects were male, 49.3 percent were female. As for ethnic identification, 86.5 percent of the subjects were identified as White, 11.0 percent as Black, .9 percent as Oriental, .1 percent as American Indian, and 1.5 percent as Other (mixed or unknown).

Every effort was made to ensure the ethnic, geographic, and socioeconomic representativeness of the sample. However, biases resulting from self-selection factors due to the voluntary nature of final subject measurements were recognized.

TABLE 1

SAMPLE POPULATION BY RACE AND SEX

RESULTS

The anthropometry and center of gravity measurements for 8,154 infants, children, and teenagers were obtained, collated, and analyzed in these two nationwide CPSC studies. For detailed information concerning the resulting data the basic reports (Snyder et al, 1975; 1977) must be referred to. The final report of the most recent study presents the measurement results for 34 measurements of infants from two weeks to 24 months old, and for 87 measurements of children and teenagers from two through 18 years old. For each measurement, tables present the summary statistics for males, females, and combined sexes in 16 age groups, including sample size (N), mean, standard deviation, minimum measurement, 5th 50th, and 95th percentiles, and maximum measurement. Also presented in tables and graphs are the bivariate relationships for selected functional measurements with stature and weight. The linear regression coefficients and constants given in these tables provide for predicting these functional measurements on individuals or groups for which only stature and weight are known.

To illustrate the results and functional nature of the measurements, results for five of the measurements are presented in the Appendix. These measurements in order of presentation are stature, weight, frontal grip reach, distance from thumb crotch to tip of the middle finger, and seated center of gravity as a percent of sitting height.

It is worth noting that since publication of the final reports several new and diverse applications of these data have been made known to us. For example, requests for these data have been made by a foreign government relative to child restraints, a major aerospace company relative to wrist dimensions, a manufacturer of medical x-ray phantoms for facial data, a U.S. assistance program for application to nutritional problems, several toy manufacturers, an automotive manufacturer relative to child dummies, a group concerned with design of school bus seats, and several other government agencies. While no single study can fulfill all the needs of all users it is the authors' hope that this body of new data made possible by the major technical advances in instrumentation of the automated computerized anthropometry system, will provide the user with basic functional, shape, and linkage data necessary for improved product safety design.

APPENDIX

APPENDIX (cont.)

REFERENCES

Alphonse R. Burdi, Donald F. Huelke, Richard G. Snyder, and George H. Lowrey, "Infants and Children in the Adult World of Automotive Safety Design: Pediatric and Anatomical Considerations for Design of Child Restraints," Journal of Biomechanics, 2(1969), 267-280.

"Requirements for Full-Sized Baby Cribs," Title 16, Chapter 2, Subchapter C, Part 1508, Federal Register, 38 (November 21, 1973), 32129-32132.

Stanley M. Garn, "Automation in Anthropometry," American Journal of Physical Anthropology, 20(1962), 387-388.

W. A. Marshall, and C. O. Carter, "Child on Mid-Parent Regression for Full Adult Stature," Annals of Human Biology, 2 (Fall 1975), 307.

Clyde L. Owings, Lawrence W. Schneider, Richard G. Snyder, and Martha L. Spencer, "A Portable System for Infant and Child Center of Gravity Measurement" Proceedings, Conference on Engineering, Medicine and Biology, (October, 1974), 375.

"Computer Controlled Anthropometry: A Portable System for Use with Infants and Children" Proceedings, Conference on Engineering, Medicine and Biology, (October, 1974), 385.

Clyde L. Owings, Richard G. Snyder, Martha L. Spencer, and Lawrence W. Schneider, "New Techniques for Infant and Child Anthropometry: Mini-Computer Controlled Anthropometry and Center of Gravity Measurements," American Journal of Physical Anthropology, 41(November, 1974) 497.

E. A. Pascarella, J. P. Foley, D. Levine, and J. R. Steward, Characteristics of Youthful Bicycle Riders in an Urbane Community and Events Occurring to Operation of Their Vehicles (Chapel Hill: University of North Carolina Highway Safety Research Center, June, 1971).

B. A. Prahl-Andersen, J. Pollman, D. J. Roaben, and K. A. Peters, "Automated Anthropometry," American Journal of Physical Anthropology, 37(1972), 151-154.

Guy G. Reiff, Leslie Kish, and Jean Harter, "Selecting a Probability Sample of School Children in the Coterminous United States," American Association for Health, Physical Education, and Recreation, Research Quarterly, 39 (May, 1968), 409-414.

Herbert M. Reynolds, Joseph W. Young, John T. McConville, and Richard G. Snyder, Development and Evaluation of Masterbody Forms for Three-Year Old and Six-Year Old Child Dummies (U.S. Department of Transportation, National Highway Traffic Safety Administration, Washington, D.C. DOT HS-801811, January, 1976.

Richard G. Snyder, Martha L. Spencer, Clyde L. Owings, Lawrence W. Schneider, and M. Anthony Schork, Selected Infant Anthropometry Crib Slat Sub-Study (Prepared for Consumer Product Safety Commission, Bethesda, Md. Final Report UM-HSRI-BI-74-4, Highway Safety Research Institute, The University of Michigan, Ann Arbor, Michigan.

Richard G. Snyder, Martha L. Spencer, Clyde L. Owings, and Peter Van Eck, Source Data of Infant and Child Measurements: Interim Data. (Prepared for Children's Hazards Division Bureau of Product Safety, Food and Drug Administration, Bethesda, Md. The University of Michigan, Ann Arbor. January, 1973; reprinted January, 1975).

Richard G. Snyder, Martha L. Spencer, Clyde L. Owings, and Lawrence W. Schneider, Physical Characteristics of Children as Related to Death and Injury for Consumer Product Safety Design. (Prepared for Consumer Product Safety Commission, Bethesda, Md. Final Report UM-HSRI-BI-75-5, Highway Safety Research Institute, The University of Michigan, Ann Arbor, Michigan. 1975 NTIS PB-242221 $7.50). Also published as, Anthropometry of U.S. Infants and Children (Society of Automotive Engineers, Inc., Warrendale, Pa. Special Publication SP-394, Paper No. 750423, February, 1975).

Richard G. Snyder, Lawrence W. Schneider, Clyde L. Owings, Herbert M. Reynolds, D. Henry Golomb, and M. Anthony Schork, Anthropometry of Infants, Children, and Youth to Age 18 for Product Safety Design (Prepared for Consumer Product Safety Commission, Bethesda, Md. Final Report UM-HSRI-77-17, Highway Safety Research Institute, The University of Michigan, Ann Arbor, Michigan. May, 1977. NTIS PB-270227 $16.25).

University of Michigan, Highway Safety Research Institute "New Methods of Infant and Child Anthropometry for Product Safety" (Produced for Consumer Product Safety Commission, Bethesda, Md. March 21, 1975. Film, 18 minutes, color).

U.S. Bureau of the Census, Statistical Abstract of the United States: 1976. (97th edition) Washington, D.C., 1976, 27.

U.S. Department of Health, Education, and Welfare, Injury Control Program, Estimates of Injuries from Consumer Products. Submitted at National Commission on Product Safety Hearing of October 21, 1969.

Joseph W. Young, John T. McConville, Herbert M. Reynolds, and Richard G. Snyder, Anthropometric Dimensions Representative of Average Three and Six-Year Old Children Sizes for the Construction of Mastermodel Body Forms (U.S. Department of Transportation, National Highway Traffic Safety Administration, and Federal Aviation Administration, Washington, D.C., Oklahoma City, Okla. DOT HS-801638. April, 1975).

----------------------------------------