skinfold thickness of youths12=17years

75
Data from the NATIONAL HEALTH SURVEY Skinfold Thickness of Youths12=17 Years UnitedStates Skinfold measurements at five anatomical sites are presented and discussed by age, sex, race, and geographic region of the country for youths 12-17 years of age in the United States, 1966-70. DHEW Publication No. (HRA) 74-1614 U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service Series 11 Number 132 Health Resources Administration National Center for Health Statistics Rockville, Md. January 1974

Upload: others

Post on 11-Dec-2021

6 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Skinfold Thickness of Youths12=17Years

Data from theNATIONAL HEALTH SURVEY

Skinfold Thicknessof Youths12=17 Years

UnitedStates

Skinfold measurements at five anatomical sites are presented

and discussed by age, sex, race, and geographic region of the

country for youths 12-17 years of age in the United States,

1966-70.

DHEW Publication No. (HRA) 74-1614

U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE

Public Health Service

Series 11Number 132

Health Resources AdministrationNational Center for Health Statistics

Rockville, Md. January 1974

Page 2: Skinfold Thickness of Youths12=17Years

Series 11 reports ~ :sent findings from the National Health ExaminationSurvey, which ol dins data through direct examination, tests, and meas-urements of samples of the U.S. population. Reports 1 through 38 relateto the adult program, Cycle I of the Health Examination Survey. Thepresent report is one of a number of reports of findings from the chil-dren and youth programs, Cycles II and III of the Health ExaminationSurvey. These latter reports from Cycles 11and III are befng publishedin Series 11 but are numbered consecutively beginning with 101. It ishoped this will guide users to the data in which they are interested.

Vital and Heakh Statistics-Series 11-No. 132

For saleby the SuperintendentofDoorrnmnts,U.S. Qover’mnentPfitfrx Omce,Washington,D.C. ‘20402- price $1.0S

Page 3: Skinfold Thickness of Youths12=17Years

NATIONAL CENTER FOR HEALTH STATISTICS

EDWARD B. PERRIN, Ph.D., Director

PHILIP S, LAWRENCE, SC.D., Deputy DirectorGAIL F. FISHER, Assistant Director for Health Statistics Development

WALT R. SIMMONS, M.A., Assistant Director for Research and Scientific Development

JAMES E. KELLY, D.D.S., Dental Advisor

EDWARD E. MINTY,, Executive Officer

ALICE HAYWOOD, Information Officer

DIVISION OF HEALTH EXAMINATION STATISTICS

ARTHUR J. McDOWELL, Director

GARRIE j, LOSEE, Deputy Director

PETER V. V. HAMILL, M.D., Medid Advisor, Children and Y6uth Progwns

HENRY W. MILLER,. Chiej Operations and Quality Control Branch

JEAN ROBERTS, Chief Medical Statistics Branch

SIDNEY AmAHml, chie~ Nuti’tion Statistics Branch

COOPERATION OF THE BUREAU OF THE CENSUS

In accordance with specifications established by the NationalHealth Survey, the Bureau of the Census, under a contractualagreement, participated in the design and selection of the sample,

and carried out the first stage of the field interviewing and certainparts of the statistical processing.

Vital and Health Statistics-Series 11 -No. 132

DHEW Publication No. (HRA) 74-1614

13braty of Congress Catabg Card Number 7$600228

Page 4: Skinfold Thickness of Youths12=17Years

CONTENTSPage

Introduction -------- -------- -------------- ------- -------------- -------

Method --------------------------- ---------------------------- --------

Results --------------- -------------- -w----- ------- -------------- -----

Age Differences -------- ..,D---- --------------------- -------------- --Sex Differences-- -------------- ------- ------- -------------- ------- -

Race Differences ---------------------------- -------------------- ----

Relationships to Data from Cycles I and H-------------------------- ---

Correlations Among Skinfolds -------- -------------- ------------------

Geographic Differences --------------------------- ------------------

Triceps Skinfold Thickness by Age and Weight -------------------------Comparison With Other Studies --------------------------- ------------

Discussion --------- ------------------ --------------------------- -----

Age and Sex Differences ----- ”--------------------- ------------------

Race Differences ----------------------- -------- ------- ------- ------Correlations Among Skinfolds ----------------------------------------

Summary ----- --------- --------------------------- --------------------

References -----------------------------------------------------------

List of Detailed Tables --------------------------- ----------------------

Appendix 1. Statistical Notes ---------------- -------- ------- ------- ----

The Survey Design -------- ------- --------------- ------- ------- ------

Some Notes on Response Rates --------------------------- ------------

Parameter and Variance Estimation ----------------------------------Standards of Reliability and Precision ------- --------------------- ----

Hypothesis Testing -------- -“----- -------------- ---.--- -w----- ------

Analysis of Correlations Among Skinfolds ------- ------- ------- ------- -Imputation ---------------------------------------------------------

1

2

336789

101010

12121415

16

17

18

5252535454545555

Page 5: Skinfold Thickness of Youths12=17Years

CONTENTS-COn.

Page

Appendix II. Demo~aphic Variables -----------------------------------Age and Sex--------------------------------------------------------Race --------------------------------------------------------------Re@on ------------------------------------------------------------

Appendix III. Techniques ofMeasurement and Quality Control -------------Inwoduction --------------------------------------------------------HES Measuring Technique --------------- ------- ---” ----------- ----.-Surveillance and Evaluation of Residual Measurement Process Error----

Monitoring Systems ----------------------------------------------Biases and Controls in Replicate Measurements ---------------------Selection of Replicate Examiners ----------------------------------

Evaluation of Residual Measurement Error in Skinfold Measurements----Results of Replicate Examinations ------------------------------------

Inter-Examiner ~fferences ---------------------------------------Intra-Examiner Differences -------- -------- -“------ -“----- ----------Conclusions -------- ---------------------- --------------------- .-

SYMBOLS

Data not available ------------------------ ---

Category not applicable ------------------- . . .

Quantity zero ---------------------------- -

Quantity more thanO but less than O.05---- 0.0

Figure doesnot meet standardsofreliabilityor precision ------------------ *

57575757

585860626262636465656567

Page 6: Skinfold Thickness of Youths12=17Years

SKINFOLD THICKNESS OF YOUTHS12-17 YEARS, UNITED STATES

Francis E. Johnston, Ph.D., Peter V. V. Harnill, M.D., M.P.H., and Stanley Lemeshow, M.S.P.H.a

INTRODUCTION

This report presents skinfold measurementsof youths 12-17 years of age in the noninstitution-alized population of the United States, as estimatedfrom the Health Examination Survey (HES) of1966-70. Findings have been analyzed by age, sex,race, and geographic region. Publication of thisreport completes a series describing and analyz-ing the quantity and distribution of subcutaneousfat in individuals and groups over the age rangeof 6-79 years.

The HES was conducted as a series of sepa-rate programs, called cycles, each of which ex-amined national probability samples of a specificage range of the noninstitutionalized population ofthe United States for a variety of information re-lated to health and development.

Cycle I in 1959-62 focused on the prevalenceof certain chronic diseases and the distribution ofvarious physical and physiological measuresamong the adult population, ages 18-79 years, asdescribed previously.1’2

During Cycle II in 1963-65, a range of fac-tors were examined in children 6-11 years of ageincluding those related to health, growth, anddevelopment.3

Cycle III, 1966-70, the program on which thefindings in this report are based, was further de-signed to obtain basic measures of growth and

%ofesmr of Anthropology, University of Pennsylvania,phi]adclpMa; Medical Advisor, Children and Youth pr0gr~s2Division of Health Examination Statistics; and forrnerly@dyt-icd Statistician, Division of Health Examination Statistics,twpectively,

development, as well as data on other health-related characteristics of the adolescent popu-lation ages 12-17 years. The survey plan, sampledesign, examination content, and operation of thissurvey program have been described in a previousreport.4

Most of the results of these three cycles areappearing in this series (Series 11) of the Vifuland HSUMZStatistics reports. Skinfold measure-ments from Cycle I (adults 18-79 years) werepublished in Series 11, No. 35;5 those from CycleII (children 6-11 years) were published in Series11, No. 120.6

For Cycle III, a national probability y sampleof 7,514 noninstitutionalized youths 12-17 years ofage in the United States was selected. Of thissample, 90 percent were examined. This nationalsample and the examined group constitute arepresentative sample of the 22.7 million nonin-stitutionalized youths 12-17 years in the UnitedStates with respect to age, sex, race, and geo-graphic region.

From the standpoint of human growth anddevelopment, the adolescent period is one of in-tense interest and extreme importance. It is &tr-ing these years that the body structure of the childis transformed into the morphology of the adult.Apart front the secondary sex characteristics,much of the sexual dimorphism observable inadults arises during adolescence and virtuallythe entire body in some way participates in thephenomenon known as the adolescent spurt, char-acterized by marked changes in the growth ratesand by rapid movement to adulthood.7 For thesereasons the battery of measurements taken inCycle II was altered and expanded for Cycle III,

1

Page 7: Skinfold Thickness of Youths12=17Years

after considerable discussion and consultation, sothat the data obtained from Cycle III would reflectthe changes in size, shape, and body compositionassociated with adolescence.

One of the most important features of humandevelopment, and one receiving more and moreattention, is the composition of the body, i.e., theconstituents of its mass, and the changes of thesecomponents. While there are many models usedto partition mass, from a quantitative aspect, intoits component tissues, depending upon the in-vestigator’s instrumentation and the problemunder consideration, the measurements of thethickness of skinfolds at specific and appropriatesites provide an acceptably accurate and demon-strably meaningful parameter for estimating thedegree of fatness of the body. 6 In addition, it isthe only approach feasible for epidemiologicalsurveys involving large sample sizes. Finally, itis basic to the evaluation of fatness/leanness inthe clinical examination of individual subjects.The use of skinfolds and the integration of thesedata with those obtained from more complex andinvolved procedures have been covered in a num-ber of publications, e.g., by Bro~ek8 and byMalina.g

This report presents the distributions ofskinfold thicknesses among youths 12-17 yearsold at five sites on the body, as measured in CycleIII of the HES, and as related to age, sex, race,and geographic region, along with certain analysesof the differences noted. The format is similar tothe report on skinfolds from Cycle II 6 to allowextrapolation and integration, and where appro-priate, data from the 6 through 11-year-old sam-ple are included to provide a more comprehensiveview of the growing years.

Statistical notes, including survey design,response rates, and reliability of the data, areincluded in appendix I. Appendix II lists pertinentdemographic variables, while appendix 111de-scribes measurement techniques and quality con-trol procedures.

METHOD

At each of 40 preselected locations through-out the United States the youths were brought tocentrally located mobile examination centers foran examination which lasted about 3 hours. Ideally,

six youths were examined in the morning and sixin the afternoon.

When the children entered the examinationcenter, their oral temperatures were taken and acursory screening for acute illness was made;if illness was detected, the youth was sent homeand examined later. The examinees changedinto shorts, cotton sweat socks, and a lightsleeveless topper, and proceeded to differentstages of the examination, each one following adifferent route. There were six different stationswhere examinations were conducted simulta-neously and the stations were exchanged. At thesestations there were examinations by a pediatri-cian, a dentist, and a psychologist, and at theothers highly trained technicians performed anumber of other examinations, including X-raysof chest and hand-wrist, hearing and vision tests,respiratory function tests and electrocardiog-raphy, an exercise test (treadmill), a battery ofbody measurements, and a grip strength test.

In Cycle I of the HES two skinfolds weremeasured. One more was added for Cycle H. Theconsiderations mentioned in the hit roduct ion war-ranted another increase for the Cycle HI exami-nation of youths and consequently two more wereadded, bringing the total to five skinfoldso

All measurements were made on the rightside of the subject, if possible, and recorded tothe nearest half-millimeter (mm.). Measure-ments were taken twice, and, where necessary,any discrepancies were resolved by means of athird measurement. In all cases a Lange skin-fold caliper was used; this instrument is designedto exert a constant pressure of 10 grams/mm. 2throughout the range of jaw openings. The pre-cision of the caliper was tested daily by check-ing it against metal standards of known widths.

Periodic training sessions were conductedby outside consultants to insure continued pro-ficiency in the measurement techniques and toobtain replicate data for the purpose of quanti-fying observer error. The results of the repli-cate examinations are presented in appendix HItogether with a description of the technique ofmeasuring skinfolds.

Each of the skinfolds measured was selectedeither because data were available from CyclesI and II and/or because each provided a signifi-cant piece of information on the dynamics of sub-

2

Page 8: Skinfold Thickness of Youths12=17Years

cutaneous fat deposition during adolescence. Therationale for selection is described below. Dia-grams of the five skinfold sites are shown infigure 111in appendix III.

The triceps skinfold was measured over the

triceps muscle at the midpoint of the upper arm.In addition to it being the site at which skinfoldmeasurements are most frequently taken, itprovides information on the quantity of extremity

fat and may be a valuable indicator of obesity.*OThe subsca@davskinfold was measured just

below the inferior angle of the right scapula. It isthe next most commonly measured skinfold and it

allows an estimate of fat on the trunk. Somestudies have suggested that it may be the skin-fold which, by itself, is most highly correlatedwith total body fat.11

The midaxillwy skinfold, taken in the mid-line as its name implies, and halfway between thenipple and the umbilicus, provides another esti-

mate of trunk fat at a site where the thickness ofthe adipose layer is minimal.

The suprailiac skinfold was added in CycleIII to permit an estimate of subcutaneous fat

deposition at what is perhaps the major “de-

pot” of the body where quite marked amounts offat accumulate in some individuals. The skinfoldwas picked up over the crest of the ilium inferiorto the midaxillary region and measured along a

natural line with a front-to-back axis.The medial calf skinfold, also added in Cycle

III, was measured to provide information on lowerlimb fat and to enable one to conclude whetherthe triceps fold was in fact a valid indicator ofboth the upper and lower extremities. In a few

individuals the skin and underlying tissues were

“stretched” so tightly that a satisfactory foldcould not be picked up at this site. No measure-ment was recorded for these subjects; however,

they amounted to only 1 percent of the total

sample measured.In all of the reports from the HES, age is ex-

pressed as the years attained at the last birthday,and the grouping for this report follows this con-vention. The mean age of each category, there-fore, approximates the midpoint of the whole

year; e.g., the 13-year-old male group consistsof a 1-year cohort whose mean age is 13.49years, while the corresponding female sampleaverages 13.48 years. The ages were validated

from birth certificates in 92 percent of the sub-jects.

“Race” was recorded as “white,” “Negro,”

and “other races. ” The white youths made up

83.98 percent of the total, the Negro youths 15.71percent, and youths of “other races” only 0.32percent. Because so few youths were classifiedas “other races, ” data from them were not ana-

lyzed separately. These data are included when“total” is used but are dropped when a white/Negro dichotomy is employed.

RESULTS

Age Differences

Tables 1-5 present the basic distributions of

the data by age and sex. All of the values werederived from the weighted sample sizes. As dis-cussed in the report of skinfolds from Cycle 11,6

the mean and its standard error are presentedmainly for the information of those investigatorswho continue to use them. However, the markedskewness of all distributions strongly argues for

the use of the median (50th percentile) as the best

measure of central tendency and for percentilesas measures of variability. These parameters areused in the analyses in this report; the meansare used only when special information can be

gained from them.The percentiles are graphed by age and sex

for each skinfold in figures 1-5 to permit a visualevaluation of age trends and sex differences.

The skewness described for the Cycle II

sample is also quite evident in these figures, the

distance from the 50th to the 95th percentile beingconsiderably greater than that from the 5th to the50th, regardless of the age, sex, or site.

Among girls, there is a general tendency for

the medians to increase from one year to the next,

indicating a steady accumulation of fat throughoutadolescence. However, there are some exceptionsto this statement which warrant pointing out. Themedian midaxillary fold increases through age 16,but the value for 17-year-olds is less. At thesuprailiac site, the median skinfold thickness for

16- and for 17-year-olds is less than for the

preceding age groups. For the 17-year-old group,the decrease is in the order of the increases notedfor earlier years and, in view of the design and

3

Page 9: Skinfold Thickness of Youths12=17Years

MALE

‘0 r30

28

26

24

22

20

18

16

14

12

10

8

6I

4t-

FEMALE Percentile

~::

:~ :~12 13 14 15 16 17 18 12 13 14 15 16 17 18

AGE IN YEARS AGE IN YEARS

Figure L Distributions by selected percentiles of the triceps skinfold of U.S. youths, by age and sex.

32-

30

28

26

24

22

20 —

16

16

14

12

10

8 —

6

4

2

o_

FEMALE Percentile

+.

@h

y:

~’$

I I I I I I o~12 13 14 15 16 17 18 12 13 14 15 18

AGE IN YEARS AGE IN YEARS

Figure 2. Distributions by selected percentiles of the subscapular skinfold of U.S. youths, by age and sex.

4

Page 10: Skinfold Thickness of Youths12=17Years

28 —FEMALE

: : ~:;’

!: ; 2“”y

: : ~’”’hu.z:

g ;; : ~’”’h ;: : ~—’’”

~

g ‘;10

5 ~ 75’6 —

8 ~ ~::

~ %6

4 —~ ‘$

4 ~ ‘$

2 — 2

0 I I I I I J o I I I I I I12 13 14 15 16 17 18 12 13 14 15 16 17 18

AGE IN YEARS AGE IN YEARS

Figure 3. Distributions by selected percentiles of midaxillary skinfold of U.S. youths, by age and sex.

FEMALE

: :2: !~-::

g; ~ ~’oth : ~ ~’”’h

y 24 — 24 —

g 22 —zg 20 —

u18 —

: : ~ 75%’<~g

: : : ~75’h ; : ~_~m,h

12 —

‘; :~; ‘: ~ ~ Z’th

4—~ 5t’ : - ~ ‘2[

2— 2—0 I I I I I 1 0 I I I I I

12 13 14I

16 16 17 18 12 13 14 15 16 17 18

AGE IN YEARS AGE IN YEARS

5

Page 11: Skinfold Thickness of Youths12=17Years

“ rMALE FEMALE

Percmtllo

!5-fc1

“l-32t-30t

2!-

o~12 13 14 15 16 17 18

AGE IN YEARS

36

34

[

32 —

30

28

26

24

22 —

20 —

18 -

16

14 -

12 —

!0 —

8 –

8 -

4 -

2 —

@h

~th

o~12 13 14 15 1-/ 18

AGE IN YEARS

Figure 5. Distributions by se)ected percentiles of medial calf skinfold of U.S. youths, by age and sex.

sample sizes of Cycle III, it is difficult to ascribeit to sampling error.

For the triceps, medial calf, and subscapu-lar skinfolds, the medians for girls show a con-sistent increase from 12 through 17 years.

On the other hand, the curves for boys showbasic differences among the five sites. The threeskinfolds which reflec~ trunk fat (subscapular,midaxillary, and suprailiac) display the samegeneral pattern of increase as noted above forgirls, though the slopes of the lines joining themedians are Iess, suggesting a sIower rate ofaccumulation of fat in boys than in girls, In con-trast with the pattern for girls, the median skin-folds for the triceps and medial calf generaIly de-crease with age, indicating a relative loss of fatin the arms and legs in boys during adolescence.

In terms of absolute values, the greatestmedians are to be found for the suprailiac and themedial calf skinfolds. The percentile distributionsof these two overIap a great deal, and it is there-fore difficult to make positive statements aboutone of the two skinfolds having greater vaIuesthan the other. However, with respect to themedians, thicker skin folds are to be found at the

medial calf among girls. The same is true inboys 12-14 years, but thereafter, as the medialcalf medians decrease, those for the suprailiacfold become greater.

The skewness in skinfold distribution was

quantified by dividing the difference between the95th and 50th percentiles by the difference betweenthe 50th and $th. b Thus a symmetrical distrilm-tion would have a value of 1.00. Skewness valuesare shown in table 6. In all cases the curves wereskewed to the right, as evidenced by values inexcess of 1.00 and they were frequently markedlyso, with values sometimes greater than 6.0.Greater skewness was noted for trunk than forlimb fat.

Sex Differences

The differences in skinfold thicknesses be-tween boys and girls are twofold. First, girls

bThis method was devised because of its simplicity, While

it is recognized that there are other measures, e.g., the con~pu-tation of the third moment about the mean, they are muchmore complex in their calculation and yield equivalent infor-mation.

6

Page 12: Skinfold Thickness of Youths12=17Years

may be observed, at any site and at any age, tohave greater median thicknesses than do boys.At the subscapular and suprailiac sites, the dif-

ferences between male and female medians showno pattern of change; at the midaxillary site, thedifferences become greater with age.

Boys and girls differ also in the curves of

sltinfold thickness and age for the triceps andmedial calf. While the medians increase with ageamong girls, there is a persistent decrease inboys from 12 through 17 years. This decrease

serves to accentuate the sexual dimorphism by 17

years. By that age the 50th percentile values forgirls are more than two times those for boys. Infact, the male medians for either skinfold fall

below the 5th percentile values for females.In contrast to the greater skinfold thicknesses

among females, there is a greater skewness to beseen, at any skinfold and at any age, among males

(see table 6). With five skinfolds and six agegroups, 30 skewness comparisons are possible;greater values are seen among males in all 30.(Note that this is a relative matter since greater

skinfold thicknesses occur among females.)

Race Differences

Tables 7-11 present the distributions of theskinfolds for whites and Negroes separately. Themedians are graphed in figures 6-10. For the

— white..—. Negro

~Fema’e.—--”~ ‘ate

ol-17 18

AGE IN YEARS

Figure 6. Median triceps skinfold of U.S. youths, by age, sex,

20

l--

— WII ihe. . . . . Negro

15

. . . . . .#

10. . . .

. ...”.”..

. ..-_.. o-....”

5

Female

Male

-112 13 14 15 16 17 1s

AGE IN YEARS

Figure 7. Median subscapular skinfold of U.S. youths, by age,sex, and race.

limb skinfolds, whites have markedly greaterthicknesses than do Negroes of the same age and

sex. Within each sex, the shapes of the curves

joining the medians for the limb skinfolds arequite similar; i.e., the 50th percentile values aregenerally decreasing in males while increasing

in females. The increases in the medial calfskin-fold from year to year are greater and moreregular in white than in Negro girls; the latter

display an irregular curve though the sample size

20 —— White-... Negro

.

E ,5n

dLLgYm~ ,0

a Female:x~ *

55 — Male

z<

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0Y

o I I I I I I12 13 14 15 16 17 18

AGE IN YEARS

Figure 8. Median midaxillary skinfold of U.S. youths, by age,sex, and race.and race.

7

Page 13: Skinfold Thickness of Youths12=17Years

20-r ——White.. . . . . Negro

~CJ

150“k:..-” ”-.................,. <,00... . . . . . . . Female,..2z ....”w . . . . . . . . . . . . . ....””~ 10

= Male

$

I. ..- . . . . . . ..ooem.......

; .............................................

o~12 13 14 15 16 17 18

AGE IN YEARS

Figure 9. Median suprailiac skinfold of U.S. youths, by age,sex, and race.

of the Negroes is smaller, which may contributeto the irregularity.

Racial differences in trunk skinfolds varywith respect to the site being discussed. For themedian subscapular fold, no systematic differ-ences are to be seen, in either sex, between whitesand Negroes. For the midaxillary, whites haveslightly greater medians than do Negroes, withmore consistency noted among boys. As far asthe suprailiac fold is concerned, racial differencesare the greatest of any of the three trunk skinfolds.

=Fema’e-“<

““”””------- ..................... ..<.00 . . . . . . . . . . . . . . . ..< .000Male

I I I I I I12 13 14 15 16 17 18

AGE IN YEARS

This difference is especially marked among boys,with the white medians being on the order of 50percent greater than those of Negro boys of thesame age group.

Skewness coefficients are presented forwhites and Negroes in table 12. As when the raceswere combined, the trunk distributions are morehighly skewed than those of limb folds. However,the magnitude of the trunk/limb dichotomy isgreater among whites.

Relationships ta Data from

Cycles 1 and II

The triceps and subscapular skinfolds weremeasured in all three cycles of the HES and hencehave now been described throughout the age rangefrom 6 through 79 years. 516In figures 11 and 12 themedians are graphed against age, males and fe-males separately, for 6-21 years. (The 21.5 yearplot represents the midpoint of those subjects18-24 years old.) The figures show the changes inmedian skinfold from 6 years of age into the adultperiod in both sexes, all races combined. The dif-ferences between the sexes in the thickness of bothskinfolds are clearly delineated and exist at allages.

For the triceps fold, the curves are, as notedabove, markedly different in shape for the twosexes, the focus of the difference centering atadolescence. From 6 years, the medians increase

17—/.” I

.. . ...

16 — Fem,k . ... . .. ...””’”... . . . .

15 — ....”...

#

2 ....”g 14 —

....”

5 ...””: 13z ...””

% ,3 _ ....”.... . . .. . . . . . . . . ..

#....”

s ,, ....”E ....”% ....”E 10 .....”

:/

9 _..””

8 —

$,, ,,, ,,, ,,, ,I I I

6780 10 11 12 13 14 15 16 17 21 22

AGE IN YEARS

Figure 10. Median medial calf skinfold of U.S. youths, by age, Figure 11. Median triceps skinfold of U.S. males and femalessex, and race. 6-21 years of age.

8

Page 14: Skinfold Thickness of Youths12=17Years

Q*~10 11 12 13 14 15 16 17 21 22

AGE IN YEAflS

Figure 12. Median subscapular skinfold of U.S. males andfemales 6-21 years of age.

in both boys and girls, boys reaching a maximumthickness in the 11 and 12 year age groups; ingirls the medians level off at 10%-12% years.Thereafter the sexual dimorphism increasesmarkedly. The decrease in median triceps thick-ness in males is such that, in the 15-17 yeargroups, the medians are less than those observedat 6-8 years. The 18-24 year group median in-dicates another increase in males from this lateadolescent “low” to a value of 12.0 mm., accord-ing to data from Cycle I, in the 25-34 year group.

Until llM years of age, the subscapular skin-folds increase in males and females in parallelfashion, though females have consistently thickermedians by about 25 percent. From 11%years on,however, the sex differences increase, with fe-males adding fat more rapidly than before, as in-ferred from changes in the medians. The appar-ent rate in males remains the same. Both sexescontinue to display increases into adulthood untilthe 5th and 6th decades.

Correlations Among Skinfolds

Tables 13 and 14 present the correlations be-tween pairs of skinfolds by age, sex, and race, aswell as for combined categories. Data from onlythose individuals on whom all skinfolds weremeasured were utilized, and the sample sizes aretherefore less than those given in tables 1-5.

For any pair of skinfolds, the correlation isquite high and is greater than 0.6 for the 450

values of both tables except in seven cases, whichwere in the 0.56-0.59 range. All values werepositive. Thus, there is a strong tendency for in-dividuals who have a greater thickness of sub-cutaneous fat at one site to be correspondinglyfatter at other sites, regardless of their age, sex,or race.

In all but a very few instances, the correla-tions among boys are higher than among girls,indicating somewhat greater independence of skin-fold thicknesses in girls from 12 through 17 years.In fact, the seven low values referred to in thepreceding paragraph all occur in girls.

With respect to race, higher correlationsgenerally are to be found among Negro boys thanwhite boys of the same age. From the age-groupedvalues of table 14 it can be seen that the v valuesamong Negro boys are higher than the correspond-ing ones for whites for all 10 skinfold pairs. Amonggirls racial differences are not clear and no pat-tern emerges.

The correlations among all possible pairs ofskinfolds were analyzed, as shown in table 15, inan attempt to discern patterns. Within each age-sex-race group, the correlations for all pairs ofskinfolds were ranked from 1 (lowest) through10 (highest). For example, among 12-year-oldmales, the lowest correlation was noted for thesubscapular/medial calf skinfolds and the highestfor the subscapular/midaxillary skinfolds. Theranks of each pair were summed for males andfor females; thus, the pair with the highest sumwould yield the highest correlations for the years12-17, as determined by relative rankings.

As one might expect, the highest correlationswere to be found with correlations of one trunkskinfold with another. Thus, among both malesand females, the subscapular/midaxillary corre-lation was the highest; in fact, this pair had thehighest ~’s for every age group in both males andfemales. The triceps skinfold correlated highlywith the medial calf, though not as well as thetrunk folds correlated with each other. The tri-ceps fold also correlated moderately well with thethree trunk sites. On the other hand, the corre-lations of the medial calf skinfold with the threetrunk folds were the lowest of all in both malesand females.

The rankings of individual skinfolds may bedetermined by summing the ranks of all of the

9

Page 15: Skinfold Thickness of Youths12=17Years

correlation coefficients in which a particular oneappears. For example, adding the ranks of allr’sinvolving the triceps fold, regardless of the otherskinfold, provides an overall indication of thestrength of the correlation of the triceps with theother four. skinfolds. Among girls, the subscap-ular skinfold was associated with the highestcorrelations with other skinfolds, while the mid-axillary ranked second and the suprailiac third.Among boys, the suprailiac fold was associatedwith the highest values, while the midaxillary andsubscapular folds clustered together in the secondspot. However, the consistency of these obser-vations is much higher among girls than amongboys.

Geographic Differences

Tables 16-20 present the skinfold distribu-tions by geographic region, age, and sex. Racialgroups are combined and the composition of eachregion reflects its own racial makeup. No realdifferences are to be observed among the fourregions except that slightly smaller medianstend to exist for children from the South. Thesedifferences, however, are almost always lessthan 1.0 mm. and may reflect the greater pro-portion of ,Negroes among the population of thatsection of the country.

Triceps Skinfold Thickness by

Age and Weight

Tables 21-35 give the distribution of the tri-ceps skinfold thickness by age for weight cate-gories from below 30 to 100 kg. or more. Suchtables were also presented for the 6 to 11-year-old children of Cycle 11.6 Their utilit y stems fromthe knowledge that a given body weight may con-tain a high or a low proportion of fat, and theevaluation of this relationship in individual chil-dren is an important feature of a clinical exami-nation. These tables allow an objective evaluationof the individual relative to the triceps skinfold,and their use is described in the report on CycleII skinfolds.6 As an example, it may be seen that90 percent of 15-year-old girls weighing 55:60 kg.(121-132 lb.) may have triceps skinfolds 9,6-24.3 mm. thick. Obviously, within any weight

category, there will be striking variation in fat-ness, The consideration of triceps fold thicknessrelative to weight, and weight relative to height~gcan help materially in the evaluation of the degreeof fatness.

Comparison With Other Studies

While it is possible to compare the HESdata to those from other studies, the significanceof such an exercise would, more often than not,be uncertain and frequently comparisons mighteven be meaningless. To compare the Cycle IIIdistributions to the results of other studies ofAmerican youths would primarily test the sam-pling variability of the latter. The Health Exami-nation Survey utilized the most sophisticatedmethods possible to insure an accurate repre-sentation of the total U.S. population and of thefour geographic regions noted above. Other studieshave sampled far more restricted groups withmethods which could only increase sampling vari-ability; hence, any differences noted would be dif-ficult to interpret.

Likewise, to compare the Cycle 111data tostudies from elsewhere in the world would be toignore the fact that the amount of body fat ishighly responsive to the environment and, again,one would be hard put to ascribe differences togenetic factors, hereditary factors, or inter-actions of the two.

On the other hand, one detailed comparisonmay be made and may prove informative eventhough the interpretations of differences may bedifficult. Scott’s study 13 of children from theLondon schools is noteworthy because of itscomprehensive nature, over 25,000 children beingmeasured. In addition, his data form one of thecore samples in the growth standards for Britishchildrenll widely used throughout the UnitedKingdom. Figures 13 and 14 present the mediansand the 10th and 90th percentiles for the HES andthe London County Council (LCC) children for thetriceps and subscapular skinfolds (the two meas-ured in the LCC survey) by age and sex.

The median triceps skinfolds are greateramong American youths of either sex than theirEnglish age peers; though these differences aresmall, they are consistent throughout the agerange 12-17 years. The reverse is, in general,

10

Page 16: Skinfold Thickness of Youths12=17Years

MALE30

26

20

16

10

5

01

30

e Health Examination Survey youths

,...., London County Council youths

25

20

Percentile‘“%.. . . . .

. . . . . . . . . . . .15

““%. . . . . . ~oth

. . . . . . . . . . . . . . . . . . . . . .

10

both

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1oth 5

I I I I I I o

FEMALE

Percentile

~oth

. . . . . . . . . . . .

. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . .

. . ./...

.................

I I I I I I12 13 14 15 16 17 18 12 13 14 15 16 17 18

AGE IN YEARS AGE IN YEARS

th Seth, and 9t)th percentiles.Figure 13. Comparison of triceps skinfold of U.S. and English youths: 10 ,

true for the subscapular skinfol~ the medians At the 10th percentile, London males displayare greater among the boys and girls of the higher values for lmth skinfolds. The same is trueLondon sample. This is in agreement with com- for the subscapular fold in females, but, in theparisons made between the London and American case of the triceps, American females have higherchildren between 6 and 11 years.6 values after the 13th year.

30

25

20

16

10

6

0

MALE30 —

_ Health Examination Survey youths.. . . . . London County Council youths I

25

20

PercentileI

FEMALE

Percentile

~’oth........................- /-=--/’’o’h ‘5..,....................................m..,,,....... I

10

.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...””

~oth

. . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...”.

, Oth

5

[

Seth...................

.....................................................,Oth

-

I I I I I I ol-12 13 14 15 16 17 1s 12 13 14 15 16 17 1s

I I I I I I

AGE IN YEARS AGE IN YEARS

Figure 14. Comparison of subscapular skinfold of U.S. and English youths: 10th, 50th, and 90th percentiles.

11

Page 17: Skinfold Thickness of Youths12=17Years

On the other hand, at the 90th percentile,greater values are to be found among Americanyouths of similar age and sex, regardless of skin-fold. These differences exceed, in some cases,3.0 mm., and indicate that greater fatness is to befound in the HES sample, compared to the Londonsample, as evidenced by the skewing of the dis-tributions.

The shapes of the curves constructed byjoining the values for similar percentiles tend tobe the same, with few differences occurring. Someof the similarities may be of particular interest;for example, the increase in the 90th percentilevalue of the subscapular fold in 13-year-old boysfrom both samples compared to adjacent values.The parallel nature of the increase in Englishand American boys suggests the operation of fac-tors related to adolescent development.

DISCUSSION

The publication of these data on subcutaneousfat distribution among U.S. children 12-17 yearsold completes data for a 74-year age range andprovides comprehensive standards for the popu-lation from 6 years onward. This represents thefirst available data of such magnitude which hasutilized sophisticated sampling techniques in-suring a national probability sample. Thus, thepopulation of the United States, as it was in the1960’s, is adequately depicted. Because the sig-nificant differences between Negroes and whitesin the amount of subcutaneous fat have now beenrecognized, black and white American childrenmay now be evaluated on standards specific fortheir own race.

Age and Sex Differences

The changes in subcutaneous fat that aredocumented in this report for the adolescent yearsas the child becomes an adult reveal significantdifferences among the sites which are related tosex. By and large, the increase in skinfold thick-ness in girls is persistent from 12 through 17years. Among white girls the medians, at 17years, for the midaxillary and suprailiac skin-folds are less than those for 16- year-olds. Inview of sample ~ize and sampling technique, thisis not likely to be fortuitous and, therefore, twopossible explanations exist. First, this may rep-

resent a true decrease; i.e., if the same girlscomprised the 16 and 17 year subgroups, thesedifferences would still exist and would be due toa decrement in growth. Such a decrement mightoccur, for example, from a conscious weightreduction, through dieting, due to peer-grouppressure. However, since th’e HES sample iscross sectional, this cannot be determined withcomplete confidence. Second, some exogenousfactors may have affected the 17- year-olds dif-ferently than the 16-year- olds, thus leading todifferences in the medians. The absence of atruly national catastrophe argues against this ex-planation.

Viewing the combined data of Cycles II andHI, it appears that the triceps and subscapularskinfolds among girls display similar growth pat-terns. There is a steady increase from 6 through17 years, except for a plateau lasting for 2 years.In the case of the triceps, this plateau occurs at10-12 years and coincides with the period of rapidgrowth in height as seen in figure 15. The plateaufor the subscapular skinfold occurs at ages 8-10years.

Among boys, as noted earlier, fundamentalintersite differences in the pattern of age changesare clear; these changes are related to other as-pects of adolescent growth. Changes in trunk skin-folds (subscapular, midaxillary, suprailiac) showan overall increase which seems to follow quitewell the observations made in Cycle II of the HES.

The decrease for boys in the triceps andmedial calf skinfolds is consistent for both sites,and the rate of change for each age “group issimilar. This phenomenon has been observed byother investigat~rs and is believed to be a gen-uine difference. For example, the findings ofJohnston and Ivialina14’15based on measurementsof tissue thicknesses on radiographs supportthese findings as do those of other investi-gators.16’17

The decrease in limb fat in boys is clearlyrelated to their growth spurts. Changes in themedian thickness of the triceps and medial calfskinfolds are closely related to the increased in-crements between mean heights associated withthe adolescent spurt.18 There is a definite rela-tionship between the period of most rapid growthin males and the decrease of triceps skinfoldthickness, which is shown in figure 15.

12

Page 18: Skinfold Thickness of Youths12=17Years

MALE

1.5

1.0 [

.1.0k‘1’

FEMALE

1.5

[

: +=’’;’nces’

.0.5

.1,0

4.0

F

3.5

3.0

2.5

2.0

1,5

1.0

0.5

0.0

...-.... ..

... ... ....... .. .... %..

......... .. ...”.....

....

:.;...

.; Mean height.. ..

. . . . . . . . . differ nce$I I I I I 1 I I I I

6789 10 11 12 13 14 15 16 17 16

AGE IN YEARSbxallld b?. IwOwrlndIlldnl awnnm

AGE IN YEARS

%OMnll by , thmwmkdm.kq svwap.

Figure 15. Comparison of pseudo velocity curve for triceps skinfold with pseudo velocity curve for height of U.S. males and females6-17 years of age.

The meaning of the decrease in the thicknessof extremity fat in boys during adolescence is notat all clear. As Tanner7 has noted, a decrease inthe thickness of the fat ring about the arm doesnot have to mean a loss of fat, for an increase hthe volume of the underlying lean tissues, with aconstant fat volume, would result in a reducedthickness of fat as it “inevitably.. would bestretched out thinner” under these conditions.The measurement of subcutaneous tissue widths,either on X-rays or using skinfold calipers, tellsnothing of changes in the size or number of adi-pose tissue cells.

Additional information may be gained from thecalcula~ion of the estimated cross-sectional areaof fat in a circle at the level of the triceps skin-fold, Since the circumference of the arm wasmeasured at this level, this parameter is easilyderived, assuming that the triceps skinfold, as adouble layer of adipose tissue, equals the diam-eter of fat at this site.

Table 36 presents the median cross-sec-tional areas of the upper arm and of the muscle-bone complex for each age group in Cycles II andIII by sex. Cross-sectional areas of fat have beenderived by subtraction of the medians. Girls dis-play steadily increasing areas of fat from 6through 15 years, paralleling the increase notedin the thickness of the triceps skinfold. The areaof fat of the 16-year-old group is essentially thesame (6 mm.2 difference) as that of the 15-year-olds, while the 17- year -olds show an increase.

The derived cross-sectional areas of fatamong boys increase consistently through age 13.For the next 2 years, i.e., ages 14 and 15, thereis a decrease, followed by a very slight increaseat 16 and a large increase at 17 years of age.

Several conclusions may be drawn from table36. In girls no fat loss occurs in the upper armsince both the triceps skinfold and the area offat increase with age. Because there is no furtherincrease in muscle-bone area after age 15, all

13

Page 19: Skinfold Thickness of Youths12=17Years

further increase in arm circumference seen intable 36 must be due to increased fat.

In boys the decrease inderived cross-sec-tional areas of fat from ages 13 through 15, cou-pled with the decrease in median triceps skinfoldthickness, is evidence for an actual loss of fat inthis anatomical region. At other ages when thetriceps fold decreases but the cross-sectionalarea of fat increases, there seems to be no ab-solute loss of fat, but instead a thinning of thering of fat due to expansion of underlying tissues.

The relative composition of the arm may beexamined by expressing the cross-sectional areaof fat as a percentage of arm area; this may beseen in figure 16, again for Cycles II and 111datacombined. The differences between boys and girlsare quite striking. Before adolescence, the rela-tive composition of the arm undergoes littlechange, although girls have a higher percentagecross- sectional area of fat. As they enter theirgrowth spurt, there is a relative decrease in fatdue to the larger increase in muscle and bone(figure 15). As the rate of growth in muscle andbone decreases, the relative fat content againbecomes greater.

The marked decrease in relative arm fat inboys occurs from 12 through 16 years, The steep-est decrease, from 13 through 15 years, is ac-companied by an absolute loss of cross-sectionalarea of fat (table 36).

Among girls, the derived cross-sectional—

“rI Female

1....

35 ............””.....

.... ............ ..+O...+ ..””’”..... ...””... ....30 ‘“””

. ...”“.....”’

: :Y15

0

6789 10 11 12 13 14 15 16 17 16

AGE IN YEARS

Figure 16. Percentage of cross-sectional area of arm comprisedof fat, by age and sex.

14

area of fat increased steadily from 6-17 years,except at age 16, when it leveled off, and themuscle-bone area increased from 6 through 15years. Thus at ages 10-12 years, when the mediantriceps fold remains unchanged, the increase inupper arm circumference is due to the expansionof the muscIe-bone mass rather than to fat. Incontrast, after 15 years, when the muscle-bonearea remains the same, the increase in arm cir-cumference is due to an increase of subcutaneousfat.

With respect to the five skinfolds, sexualdimorphism during adolescence is greatest for thetriceps and the medial calf. The ratio of the me-dianfor females to that of males at 17 years is asfollows:

Suprailiac: 1.3Midaxillary: 1.5Subscapular: 1.6

Triceps: 2.1Medial calfi 2.2

The female/male ratio for extremity skin-folds increases steadily during the adolescentyears. The relative sex difference in extremityfat is about the same for the triceps as for themedial calf, which is contrary to the suggestionsof Tanner, 7 although the data on which he hasbased them may be subject to sampling error.

Race Differences

Differences between whites and Negroes re-ported for the Cycle 11sample6 persist through-out the ages of 12-17 years: whites have greatermedians for the triceps, when compared to Negrochildren of the same sex and age, but not for thesubscapular skinfold. On the other hand the mid-axillary skinfold medians are slightly greater inwhite girls at ages 12-17 years, a differencenot observed in the 6-11 year old group.

The median medial calf skinfold changesfrom 12 through 17 years in a manner similar tothat for the triceps in all four race/sex groups,which is as expected.

The median thicknesses of the suprailiacskinfold do not conform to the pattern of the sub-scapular or midaxillary folds, in that whites havemarkedly greater values than do Negroes. How-ever, Negro girls have greater medians than dowhite boys of the same age, so that the influenceof sex-related factors is still strong, Apparently

the subcutaneous fat on the trunk displays a gra-dient as one moves from the back down andaround to the crest of the ilium with respect to

Page 20: Skinfold Thickness of Youths12=17Years

race; greater differences between the medians areassociated with more inferior skinfolds on thetrunk.

In the report on skinfolds from the HESCycle 11,6 it was suggested that extfemity fat(as indicated by the thickness of the triceps skin-fold) differed in whites and Negroes because of thecombined effects of racial and environmental fac-tors, while trunk fat (as indicated by the mid-axillary and subscapular folds) differed because ofthe effects of environmental factors alone. Thiswas based on the observation that the racial dif-

“ferences in triceps fold thickness of children 6-11years were of two degrees: the medians weregreater among whites than Negroes and there wasalso greater skewness among whites, the compari-sons being made to groups of the same age and sex.At the same time, racial differences in the twotrunk skinfolds existed dy in the degree of skew-ness, whites showing the greater distortion of thecurves. It was suggested that environmental fac-tors would tend to skew the curves and thereforeenvironmental differences would result in differ-entials in skewness. Racial (i.e., hereditary) dif -ferences would result in a shift of the distribu-tion with corresponding differences in the me-dians.

The data from Cycle III reinforce these con-clusions, and they are combined with those fromCycle II in table 37, which presents the results ofcomparing the differences in the degree of skew-ness (given in table 12) between Negroes andwhites at each age and for all five skinfolds.Greater skewness is seen in whites for the sub-scapular and midaxillary skinfolds in 31 of 46comparisons which differed. Assuming, as before,that skew results from environmental pressures,and since the major differences between the racesin the skinfolds at these two sites are in the skew-ness and not in the medians, it is concluded thatNegro/white differences at these sites reflectenvironmental factors.

Again with regard to skewness, the reverseis true for the limb skinfolds. Greater asymmetryis seen in the triceps and medial calf skinfoldsamong Negroes h“ 26 of 35 comparisons. On theother hand, in both sexes, the median triceps andmedial calf skinfolds are consistently greater itIwhites. The hereditary factors, which presumablyaffect the medians, seem to work contrary to,and to be more important than, environmentalfactors.

The suprailiac skinfold distributions seemmore similar, in terms of asymmetry, to thoseof the limbs than to those of the trunk, and onewould conclude that the greater median valuesamong whites therefore reflect hereditary mech-anisms leading to fat deposition around the pel-vic girdle.

Table 37 also shows the comparisons ofmales and females, races combined. For everyskinfold, greater skewness is noted among malesmore often than among females. If the greaterphysical susceptibility to environmental factors ofmales is considered, support is given to the as-sumption that skewness of skinfold distributionsis produced by environmental influences.

Finally, the five skinfolds may be ranked,races and sexes combined, by their overallskewness. This can be done by averaging allskewness coefficients for a particular skinfold,resulting in the following:

Midaxillafi: 5.37Subscapular: 4.96

Suprailiac: 4.01Triceps: 3.07

Medialcalf: 2.81These values were computed only for 12-17 yearolds, since the suprailiac and medial calf skin-folds were not measured in Cycle II. In Cycle II,the three skinfolds that were measured ranked asabove.

The above ranking of the five skinfolds, interms of their skewness, was consistent for allfour sex/race groups with the following excep-tions: among Negro males, the suprailiac skin-fold yielded the highest average skewness and,among Negro females, there was a reversal ofthe midaxillary and subscapular skinfolds. All ofthis suggests that trunk fat, particularly that inthe subscapular and midaxillary regions, is moreresponsive to the environment than limb fat andthat either of these two skinfolds may be a bettermeasure of adiposity in nutritional surveys thanis the triceps skinfold.

Correlations Among Skinfolds

The high intercorrelations among all skin-fold pairs suggests that the thickness of the adi-

pose layer in an individual at one site is stronglydependent upon the thicknesses at other sites. Thegenerally high values for the midaxillary andsubscapular folds suggests that either may be thebest single skinfold during adolescent years touse in estimating body fat, although it is clearly

15

Page 21: Skinfold Thickness of Youths12=17Years

better to measure othet-s as well, if at all pos-sible. The subscapldar and triceps, the two foldsmost frequently measured, probably pro-ride ac-ceptable estimates of body fat. The nearness ofthe midaxillary to the subscapular site suggeststhat its addition would not add much more infor-mation. (Their correlations are also consistentlythe highest.) The difficulty in measuring themedial calf skinfold and the relatively low ?’values argue against a high priority for its in-clusion.

SUMMARYThis report presents and analyzes the dis-

tribution of skinfold measurements in UnitedStates youths, 12-17 years of age, as estimatedfrom Cycle III of the Health Examination Survey.The thickness of five skinfolds was measured—triceps, subscapular, midaxillary, suprailiac, andmedial calf—and the results analyzed by age, sex,race, geographic region, and in relation to youthsfrom outside the United States. As appropriate,data from Cycle H have been included to presenta picture of changes in body fatness from 6 through17 years.

At. every site, girls display greater skinfoldthicknesses than do boys of the same age. Thesedifferences exist at all percentiles and for bothNegroes and whites, similar to those found inCycles I (18-79 years) and II (6-II years).

In both males and females, the skinfolds ofthe trunk (subscapular, midaxillary, suprailiac)increase in thickness with age from 12 through17 years. In girls the median triceps skinfold,after leveling off at ages 10-12 years, continuesto increase through 17, while in boys both thetriceps and the medial calf folds decrease steadilyfrom 12 through 16 years of age.

Since these patterns of change in the tricepsskinfold occur during the period of rapid ado-lescent growth, they were also examined by de-riving the cross- sectional area of fat at the levelof the triceps skinfold for each individual from6-17 years. In addition? the underlying muscle-bone mass was calculated from the upper armcircumference and triceps thickness.

The leveling off of the median triceps skin-fold in females at 10-12 years, seen in figure 11,represents the counterpart to the reduction inlimb skinfold thickness among males. However,the magnitude of the change is not at all com-parable to that observed in males. Thus, it is

16

clear that the increase in upper arm circumfer-ence among females 10-12 years is due to anincrease of muscle and bone.

In boys, the area of fat at the level of the tri-ceps fold actually decreased from 13 through 15years of age, then increased from 16 throdgh 17years. This decrease in cross-sectional area offat, along with the sharp decrease in triceps skin-fold thickness at that time, indicates an actualloss of arm fat at ages 13-15 years, which cor-responds to the period of most rapid growth forthese boys. It appears that the decrease intriceps fold thickness at other ages is due, not tofat loss, but to the thinning of the fat ring causedby the growth of the underlying muscle and bone.

Racial comparisons reveal greater medianskinfold thicknesses in whites than in Negroes ofcorresponding age and sex for the triceps, medialcalf, and suprailiac folds, while no consi stent dif -ferences were noted for the subscapular and mid-axillary folds.

The greater skewness to the right of the sub-scapular and midaxillary distributions in whites,considered in conjunction with the similarity ofthe medians for whites and Negroes, suggeststhat racial differences at these sites resultprimarily from the operation of environmentalfactors. At the other sites, both hereditary andenvironmental factors seem to be operating.

The generally high correlations betweenpairs of skinfolds indicate a tendency through-out the entire body to fatness or lean,less in theindividual. However, those correlations in whichone of the three trunk folds was involved tendedto be the highest, while those involving the medialcalf were the lowest.

It is concluded that the measurement of thetriceps and subscapular skinfolds, being the onesmost frequently taken, offer acceptable estimatesof body fat, although others may be desirable forspecific purposes. The midaxillary fold, due toits high correlation with the subscapular, yieldsno additional information. The high correlationof the suprailiac with the subscapular, and itsrelatively high error of measurement, suggestthat its use be reserved for specific purposes.The high correlation of the medial calf with thetriceps, its low correlation with all others, andits high error of measurement, argue againstits use, as a general rule. The apparent respon-siveness of the subscapular fold to environ-mental factors, the acceptably lower errors of

Page 22: Skinfold Thickness of Youths12=17Years

measurement, and its high correlation with all ment in the study of adiposity caused by dietary

others, lead to the conclusion that it is prob- intake.

ably the single most useful skinfold measure-

REFERENCES

1Nation~ Center for Health Statistics: Plan and initial

program of the Health Examination Survey. Vital and Health

Statistics. PHS Pub. No. 1000-Series l-No. 4. Public HealthService. Washington. U.S. Government Printing Office, July,1965.

2National Center for Health Statistics: Cycle I of theHealth Examination Survey: sample and response, United

States, 1960-1962. Vital and Health Statistics. PHS Pub. No.1000-Series 1l-No. 1. Public Health Service. Washington. U.S.Government Printing Office, Apr. 1964.

3National Center for Health Statistics: Plan, operation,

and response results of a program of children’s examinations.Vikd and Health Statistics. PHS Pub. No. 1000-Series l-No. 5.Public Health Service. Washington. U.S. Government Printing

Office, Oct. 1967.4Nationd Center for Health Statistics: Plan and operation

of a Health Examination Survey of youths 12-I 7 years of age.Vitaf and Health Statistics. PHS Pub. No. 1000-Series l-No. 8.Public Health Service. Washington. U.S. Government PrintingOffice, Sept. 1969.

5 Nation~ Center for Health Statistics: Skinfolds, body

girths, biachromial diameter, and selected anthropometric in-dices of adults, United States, 1960-1962, Vital and Health

S&stistics. PHS Pub. No. 1000-Sesies 1 l-No. 35. Public HealthService. Washington. U.S. Government Printing Office, Feb.1970.

6Nation~ Center for Health Statistics: Skinfold thickness

of children 6-11 years, United States. Vital and Health Statis-

tics. Series 11-No. 120. DHEW Pub. No. (HSM) 73-1602.Health Services and Mental Health Administration. Washington.U.S. Government Printing Office, Oct. 1972.

7Tanner, J, M,: Growth at Adolescence, ed. 2. oxford.

Blackwell Scientific Publications, 1962.8 Brozek, J.: Quantitative description of body composit-

ion: physical anthropology’s “fourth” dimension. Curr. An-

thropol, 4:3-40, 1963.9Mdina, R. M.: Qumtification of fat, muscle, and bone in

man. Clin. orthop. 65:9-38, 1969.10seltzer, C. C., Goldman, R. F., and Mayer, J.: The tri-

ceps skinfold as a predictive measure of body density and bodyfat in obese adolescent girls. Pediatrics 36:212-218,1965.

1 lTanner, J. M.: Standards for subcutaneous fat in Britishchddren, Percentiles for thickness of skinfolds over triceps andbelow scapula. Br. Med.]. 1:446450,1962.

~2National Center for Health Statistics: Height andweight of children, United States. Vital and Heakh Statistics.PHS Pub. No. 1000-Series 1l-No. 104. Public Health Service.

Washington. U.S. Government Printing Office, Sept. 1970.

JS.Scott, J. S.: Report on the Heights and Weights (and

other Measurements) of School Pupils in the County of Loti-

don in 1959. London County Councfl, 1959.14Johnston, F. E., and M&a, R. M.: Age changes in the

composition of the upper arm in Philadelphia children. Hum.

Biof. 38:1-21.1966.

15M~~a, R. M., and Johnston, F. E.: Relations between

bone, muscle, and fat widths in the upper arms and calves of

boys and girls studied cross-sectionally at ages 6 to 16 years.Hum. Biol. 39:211-223, 1967.

~6Hea1d, F. p., Hint, E. E., Schwmtz, & Cook, C. D.,

Elliot, O., and Vajda, B.: Measuses of body fat and hydrationin adolescent boys. Pediatn”cs 31:226-239, 1963.

17Tanner, J. M.: Radiogs-aphic studies of body composi-

tion in children and adults, in J. Bro%ek, cd., Human Body

Composition, Approaches and Applications. London. Per-ganmn Press, 1965. pp. 211-236.

18 Nation~ Center for Health Statistics: Height =d

weight of youths 12-17 years, United States. Vital and Health

Statistics. Series 1 l-No. 124. DHEW Pub. No. (HSM) 73-1606.Health Services and Mental Health Administration. Washington.U.S. Government Printing Office, Jan. 1973.

19National Center for Health Statistics: Sample designand estimation procedures for a National-I-M&h ExaminationSurvey of children. Vital and Health Statistics. PHS Pub. No.1000-Series 2-No. 43. Public Health Service. Washington. U23.Government Printing Office, Aug. 1971.

20National Center for Health Statistics: Origin, program,

and operation of the U.S. National Health Survey. Vital and

Health Statistics. PHS Pub. No. 1000-Series l,No. 1. Public

Health Service. Washington. U.S. Government Printing Office,

Apr. 1965.21National Center for Health Statistics: Replication: An

approach to the analysis of data from complex surveYs. vitaland Health Statistics. PHS Pub. No. 1000-Series 2-No. 14. Pub-

lic Health Service. Washington. U.S. Government Printing Of-

fice, Apr. 1966.22Nation~ Center for Health Statistics: A Regression

Scheme as a Solution to the Problem of Imputation, by S.Lemeshow and T. Drizd, Technical Information Notes, No. 49.Division of Health Examination Statistics, Rockville, Md. Jan.31, 1973. Mimeographed.

23Nation~ Center for Health Statistics: Quality control in

a National Health Examination Survey. Vital and Health Statis-

tics. Series 2-No. 44. DHEW Pub. No. (HSM) 72-1023. HealthServices and Mental Health Administration. Washington. U.S.Government Printing Office, Feb. 1972.

17

Page 23: Skinfold Thickness of Youths12=17Years

LIST OF DETAILED TABLES

Table 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Triceps skinfold of youths by sex and age at last birthday: sample size, esti-mated population size, mean, standard deviations standard error of the mean, andselected percentiles, United States, 1966-70------------------------------------

Subscapular skinfold of youths by sex and age at last birthday: sample size, es-timated population size, mean, standard deviation, standard error of the meansand selected percentiles, United States, 1966-70--------------------------------

Midaxillary skinfold of youths by sex and age at last birthday: sample size, es-timated population size, mean, standard deviation, standard error of the mean,and selected percentiles, United States, 1966-70--------------------------------

Suprailiac skinfold of youths by sex and age at last birthday: sample size, es-timated population size, mean, standard deviation, standard error of the mean,and selected percentiles, United States, 1966-70--------------------------------

Medial calf skinfold of youths by sex and age at last birthday: sample size, es-timated population size, mean, standard deviation, standard error of the mean,and selected percentiles, United States, 1966-70--------------------------------

Skewness coefficients(difference between the 50th and 95th percentile divided bythe difference between the 5th and 50th percentile) by skinfold site, sex, andage at last birthday of youths: United States, 1966-70--------------------------

Triceps skinfold of youths by race, sax, and age at last birthday: sample size,estimated population size, mean, standard deviation, standard error of the mean,and selected percentiles, United States, 1966-70-------------------------------

Subscapular skinfold of youths by race, sex, and age at last birthday: samplesize, estimated population size, mean, standard deviation, standard error of themean, and selected percentiles, United States, 1966-70--------------------------

Midaxillary skinfold of youths by race, sex, and age at last birthday: samplesize, estimated population size, mean, standard deviation, standard error of themean, and selected percentiles, United States, 1966-70--------------------------

Suprailiac skinfold of youths by race,sex,and age at last birthday: sample size,estimated population size, mean, standard deviation, standard error of the mean,and selected percentiles, United States, 1966-70--------------------------------

Medial calf skinfold of youths by race, sex, and age at last birthday: samplesize, estimated population size, mean, standard deviation, standard error of themean, and selected percentiles, United States, 1966-70--------------------------

Skewness coefficients (difference between the 50th and 95th percentile dividedby the difference between the 5th and 50th percentile) by skinfold site, race,sex, and age at last birthday of youths: united States, 1966-70-----------------

Sample size and correlation coefficients for all possible pairs of skinfolds ofyouths by race and sex: United States, 1966-70----------------------------------

Sample size and correlation coefficients for all possible pairs of skinfolds ofyouths by race, sex, and age at last birthday: United States, 1966-70-----------

Rankings of correlation coefficients for all possible pairs of skinfolds of youthswith each race, sex, and age group: United States, 1966-70----------------------

Triceps skinfold of youths by geographic region, sex, and age at last birthday:sample size, estimated population size, mean, standard deviation, standard errorof the mean, and selected percentiles, United States, 1966-70-------------------

Subscapular skinfold of youths by geographic region, sex, and age at last birth-day: sample size, estimated population size, mean, standard deviation, standarderror of the mean, and selected percentiles, United States, 1966-70-------------

Midaxillary skinfold of youths by geographic region, sex, and age at last birth-day: sample size, estimated population size, mean, standard deviation, standarderror of the mean, and selected percentiles, United States, 1966-70-------------

Suprailiac skinfold of youths by geographic region, sex, and age at last birth-day: sample size, estimated population size, mean, standard deviation, standarderror of the mean, and selected percentiles, United States, 1966-70-------------

Medial calf skinfold of youths by geographic region, sex, and age at last birth-day: sample size, estimated population size, mean, standard deviation, standarderror of the mean, and selected percentiles, United States, 1966-70-------------

Page

20

20

21

21

22

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

18

Page 24: Skinfold Thickness of Youths12=17Years

LIST OF DETAILEDTAELES-Con.

Table 21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

Triceps skinfold of youths 12-17 years of age weighing less than 30 kilograms,by sex and age at last birthday: sample size, estimated population size, mean,standard deviation, standard error of the mean, and selected percentiles, UnitedStates, l966-7O----------------------------------L------------------------------

Triceps skinfold of youths 12-17 years of age weighing 30-34.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean, and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 35-39.9 kilograms, by sexand age at last birthday: sample size, estimated population size, ,mean, standard13d~;t#m, standard error of the mean, and selected percentiles, United States,

---------------------------------------------------------..---------------

Triceps skinfold of youths 12-17 years of age weighing 40-44.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean , and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 45-49.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean , and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Tricepsskinfoldof youths12-17 years of age weighing50-54.9kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean, and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 55-59.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean1966-70

, and selected percentiles, United States,--------------------------------------------.----------------------------

Triceps skinfold of youths 12-17 years of age weighing 60-64.9 kilograms, by sexand age at last birthday: sample size,deviation,

estimated population size, mean, standardstandard error of the mean, and selected percentiles, United States.j

1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 65-69.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean , and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 70-74.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean, and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps,skinfold of youths 12-17 years of age weighing 75-79.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean, and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 80-84.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean, and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 85-89.9 kilograms, by sezand age at last birthday: sample sizedeviation, standard error of the mean

, estimated population size, mean, standard, and selected percentiles, United States,

1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 90-99.9 kilograms, by sexand age at last birthday: sample size, estimated population size, mean, standarddeviation, standard error of the mean, and selected percentiles, United States,1966-70-------------------------------------------------------------------------

Triceps skinfold of youths 12-17 years of age weighing 100 kilograms or more, bysex and age at last birthday: sample size,estimated population size,mean, stand-ard deviation,standard error of the mean,and selected percentiles,United States,1966-70-------------------------------------------------------------------------

Area of arm at triceps, estimated area of muscle and bone, and estimated area offat for males and females 6-17 years of age: United States, 1963-70-------------

Comparison of skewness ratios between races and sexes for skinfolds of childrenand youths: United States, 1963-70----------------------------------------------

Page

37

37

38

38

39

40

41

42

43

44

45

46

47

48

49

50

51

19

Page 25: Skinfold Thickness of Youths12=17Years

Table 1. Tricepsskinfoldof youthsby sex and age at last birthday:samplesize, estimatedulation size,mean, standarddeviation, standarderror of the mean, and selectedpercent%;United States,1966-70

Percentile

5th 10th 25th 50th 75th 90th 95thI1sNSex and age n

Male In millimeters

10.710.59.5;.:

9:0

12.913.614.815.816.516.5

5.835.855,605.11;.:;.

5.826.076.136.606.916.24

0.270.250>.270.250.280.24

0.290.280.230.370.410.41

::;4.24.34.24.1

::;7.17.4

U

5.5

2:;4.74.74.6

;.;

8:3

;::9.6

13.3 19.612.9 19.312.0 17.110.7 15.811.1 15.811.4 15.6

23.322.620.820.720.220.5

25.225.526.729.429.729.1

12 years-----------------13 years-----------------14 years-----------------15 years-----------------16 years-----------------17 years-----------------

643626617613555489

547582586502535468

2,0322,006;:;$:

1,8331,764

1,9701,9461,9011,8481,7871,743

6.8

::;5.85.85.5

::;10.711.611.812.1

9.4;.:

7:67.57.5

11.812.414.014.815.615.8

Female

16.0 22.217.1 22.818.5 23.3;;.: ;:.;

20:5 25:0

12 years-----------------13 years-----------------14 years-----------------15 years-----------------16 years-----------------17 years-----------------

NOTE: ~= sample size; N=estimated number of youths in populationin thousands; ~-mean;S = standard deviation; S2= standard error of the mean.

Table 2. Subscapularskinfoldof youthsby sex and age at last birthday: samplesize, estimatedpopulationsize,mean, standarddeviation,standarderror of the meanjandselectedpercentiles,United States,1966-70

Sex and age

?l

543526;18i13556i87

;46582586501536i69

N

2,0322,0061,9511,9001,8361,758

1,9681,9461,9011,8451,7891,746

Percentile

z s SE

5th 10th 25th 50th 75th 90th 95th

In millimeters

121314151617

121314151617

7.38.18.18.38.59.3

1;:;11.512.413.113.3

5.446.185.625.424.925.16

6.226.846.727.237.567.72

0.220.220.290.220.220.24

0.270.360.250.320.390.49

t:4.24.55.05.1

4.55.05.5

::;6.2

4.14.34.65.15.35.6

5.15.36.26.46.86.7

4.75.15.45.76.16.5

?;

;:?8.18.2

5.76.26.56.87.27.7

U

J::10.811.3

13.716.114.213.713.816.4

18.820.920.324.024.423.0

20.822.120.420.419.521.4

23.825.625.828.930.230.6

years-----------------years-----------------years-----------------years-----------------years-----------------years-----------------

7.68.58.58.89.210.6

11.712.713.715.115.717.0

Female

years-----------------years-----------------years-----------------years-----------------years-----------------years-----------------

NOTE: n= samplesize; ~= estimatednumber of youths in populationin thousands; ~=mean;S = standard deviation; SE =standard error of the mean.

20

Page 26: Skinfold Thickness of Youths12=17Years

Table 3. Midaxillary skinfold of youths by sex and age at last birthday: sample size, estimatedDowlation size, mean, standard deviation, standard error of the meaqand selected percentiles,United States, 1966-70

Sex and age n N IEs90th ]95th

Male In millimeters

121314151617

121314151617

643626617613556488

546582586502536469

2,0322,0061,9471,9001,8361,762

1,9671,9461,9011,8471,7891,746

4.945.42.5.725.554.825.71

5.926.296.177.006.946,91

0.200.180.300.230.240.30

0.260.330.220.340.340.48

4.74.95.15.45.55.7

6.7

:::8.5

:::

6.66.96.87.37.37.9

10.611.312.613.213.613.7

13.413.513.612.512.414.8

17.118.219.521.921.420.7

19.119.119.317.917.422.2

23.323.422.826.827.724.9

years-----------------years-----------------years-----------------years-----------------years-----------------years-----------------

6.3

H

::;7.6

8.6

1%;10.911.211.0 I

3.1 3.23.1 3.43.2 ;.;3.43.4 3:83.6 4.1

4.2::: 4.54.4 4.94.5 5.2

5.4::; 5.4

N4.24.44.54.6

5.15.66.36.46.86.6

J?emale

years-----------------years-----------------years-----------------years-----------------years-----------------years-----------------

NOTE: n =sample size; N=estimated number of youths in population in thousands; ~=mean;S =standard deviation; ~ = standard error of the mean.

Table 4. Suprailiac skinfold of youths by sex and age at last birthday: sample size,Modulation size. mean, standard deviation,

estimatedstandard error of the mean,and selected percentiles,

Un~ted States, 1966-70

Sex and age x

11.812.212.112.612.613.9

13,914.715.516.316.215.9

NPercentile

5th 10th 25th 50th 75th 90th 95th

n J-S SF

In millimetersMale

640626618613555487

545582586502536468

2,0272,0061,9511,9001,8341,757

1,9631,9461,9011,8471,7891,742

9.149.499.399.038.529.83

8.368.227.758.428.128.64

0.330.350.430.370.470.55

0.370.370.340.500.490.54

:::4.2

::;5.0

4.7;.:

6:36.76.1

4.4

2:?5.35.55.5

5.66.47.3

:::7.4

5.7:.:

6:87.17.2

u10.110.310.610.2

8.38.78.89.4

1::?

1.2.213.114.114.814.713.8

15.315.314.615.116.217.6

18.718.820.321.420.320.7

26.627.626.426.425.429.6

25.526.226.629.328.328.3

31.633.833.432.332.338.1

30.832.431.2:$:

32:8

years-----------------years-----------------years-----------------years-----------------years-----------------years-----------------

Female

years-----------------years-----------------years-----------------years-----------------years-----------------years-----------------

NOTE: n =sample size; ~= estimated number of youths in population in thousands; ~=mean;S = standard deviation; “Sz= standard error of the mean.

21

Page 27: Skinfold Thickness of Youths12=17Years

Table 5. Medial calf skinfold of youths by sex and age at last birthday: sample size, estimstedpopulation size, mean, standard deviationUnited States, 1966-70

, standard error of the mean,and selected percentiles,

Sex and age

12 years-----------------13 years-----------------14 years-----------------15 years-----------------16 years-----------------17 years-----------------

Female

12 years-----------------13 years-----------------14 years-----------------15 years-----------------16 years-----------------17 years-----------------

n

639620612609553487

53657957949753046o

iv

2,0211,9801,9341,8881,8281,758

1,9331,9361,8761,8301,7681,710

Percentile

z s L$5th 10th 25th 50th 75eh 90th 95th

In millimeters

12.012.311.310.710.09.9

15.316.617.818.719.019.7

6.877.076.986.415.816.18

6.767.177.017.757.287.70

0.280.330.330.310.290.29

0.270.310.350.510.340.47

4.85.04.74.64.14.0

7.37.58.78.89.49.4

5.45.75.3

M4.5

8.5

1::;10.411.111.0

;.;

6:76.76.26.0

LO.8L1.7L2.6L3.8L4.OL4.1

10.510.69.69.28.48.5

14.115.516.817.718.518.9

15.615.614.013.112.512.2

18.721.322.522.823.424.0

21.522.420.319.418.718.5

24.526.527.928.828.630.6

26.426.925.424.422.322.6

30.231.230.634.531.734.4

NOTE: 7Z= sample size; N= estimated number of youths in population in thousands; ~=mean;s = standard deviation; S% = standard error of the mean.

Table 6. Skewness coefficients (difference between the 50th and 95th percentile divided by thedifference between the 5th and 50th percentile) by skinfold site, sex, and age at last birthdayof youths: United States, 1966-70

Sex and age

12

:;151617

121314151617

years--------------------------years--------------------------years--------------------------years--------------------------years--------------------------years--------------------------

Female

years--------------------------years--------------------------years--------------------------years--------------------------years--------------------------years--------------------------

Triceps

3.182.943.603.983.873.83

2.362.181.831.971.811.70

Subscapular

7.587.495.955.835.515.25

4.884.89‘3.854.184.253.79

Skinfold

tidaxillary

8.608.047.586.255.607.84

5.464.683.564.554.063.86

Wprailiac

5.155.445.404.874.455.00

2.49;.::

2:282.222.47

fedial calf

2.732.913.203.273.243.19

2.361.971.691.871.471.63

22

Page 28: Skinfold Thickness of Youths12=17Years

Table 7. Triceps skinfold of youths by race, sex, and age at last birthday: sample size, esti-mated population size, mean. standard deviation, standard error of the mean, and selected per-centilksi United States,1966-70

Race, sex, and age

121314151617

12

::

::17

121314151617

H14151617

m

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

NEGRO

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

n

540542526525495417

455490484424440392

1018088

H69

88

J:739374

N

1,7471,7291,683L,6461,5911,528

1,6851,6671,6331,5921,5401,500

280262256241231225

272275266235243237

IX5

11.010.89.89.39.19.2

13.013.815.116.116.816.8

8.78.5

n7.17.7

12.212.013.614.114.914.5

5.825.815.595.295.115.$7

5.785.876.086.656.726.24

5.195.825.393.083.265.11

6.166.956.355.887.875.85

0.280.270.280.270.310.25

0.340.310.260.370.400.43

0.590.650.690.340.450.59

0.660.610.560.641.541.01

5.24.84.34.34.24.1

6.16.67.37.58.18.5

3.83.63.6

:::4.2

6.06.05.57.27.07.1

In millimeters

5.7

;::4.85.04.5

7.1

::;8.8

1;::

4.74.24.14.24.34.6

6.56.36.78.27.67.6

7.2

::;

:::5.6

9.2

1;:;12.012.312.4

5.6

M5.14.95.3

7.7

1;:;10.29.610.7

9.7

::;7.87.67.7

12.012.714.215.116.016.3

7.4

:::6.4

H

10.69.812.512.813.113.4

13.613.412.511.211.611.6

16.017.218.720.021.220.8

10.310.38.4

u8.5

16.213.617.317.917.818.2

19.819.717.416.416.515.8

22.122.723.525.425.525.3

15.315.614.210.611.714.0

22.523.922.122.726.423.3

23.222.621.221.320.520.7

25.125.426.829.929.129.5

23.225.219.214.712.815.8

25.627.224.625.831.525.8

NOTE: n = sample size; l/= estimated number of youths in population in thousands; ~= mean;s G standard deviation; S% = standard error of the mean.

23

Page 29: Skinfold Thickness of Youths12=17Years

Table 8. Subscapular skinfold of youths by race sex, end a e at’last birthday: sample size, es-timated population size, mean, standard deviat$on, standar/?error of the mean, end selected per-centiles, United States,1966-70

Race, sex, end age

121314151617

12

;:151617

121314151617

121314151617

WHITE

~

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

NEGRO

&

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

n

540542527525496415

454490484423g;

101808884

z

H101

‘E74

N

1,7471,7291,6861,6461,5941,522

1,6821,6671,6331,5881,5421,502

280262256241231225

272275266235243237

]

z

7.48.28.28.5

:::

L;::1.1.6L2.6L2.9L3.2

6.37.5

;:;

;::

10.410.511.3L1.913.813.4

1s

5.466.315.675.675.135.04

6..166.926.807.417.367.79

3.925.305.292.893.145.98

6.696.406.256.048.727.27

).223.24).303.243.263.26

Q.34Q.41CI.260.360.420.52

0.440.540.780.320.260.69

0.910.510.49;.;;

1:05

Percentile

jth 10th 25th 50th 75th 90th 95th

2:M5.15.1

::$5.55.86.26.2

4.14.24.14.44.35.2

4.6;.;

6:26.16.9

In millimeters

4.1~.:

5:15.35.6

5.0

u6.46.86.6

4.24.44.44.84.75.5

5.35.76.46.77.17.3

4.75.15.4

i?:;6.5

5.86.57.4

:::8.2

4.75.15.25.45.66.6

6.67.0

:::

8:3

5.76.26.56.97.27.8

::;

L;:;10.8L1.2

5.56.26.36.47.37.6

8.48.6

1::210.512.3

w!3:;1::;

11.612.813.815.315.517.0

6,57.6

n

%:

11.811.613.613.,416.416.0

14.016.514.614.815.017.1

18.621.320.424.424.023.4

1?:210.5

1::;11.9

20.518.718.121.526.622.1

21,122.320.621.320.522.0

23.725.526.330.029.831.0

16.517.215.311.712.320.7

24.630.025.126.032.830.5

NOTE: n= sample size; N= estimated number of youths in population in thousands; ~=mean;S.= standard deviation; SE= standard error of the mean.

24

Page 30: Skinfold Thickness of Youths12=17Years

Table 9. Midaxillary skinfold of youths by race sex, and a e at last birthday: sample size, es-8ti.matedpopulation size, mean, standard deviation, standar errorof the mean, and selected per-

centiles,UnitedStates,1966-70

Race, sex, and age

W-E

Male

years---------------years---------------years---------------years---------------years---------------yeare---------------

Female

years---------------years---------------years---------------yeara---------------years---------------years---------------

NEGRO

Male

years---------------years---------------veals---------------

15 ~ears---------------16 years---------------17 yeara--------------

Female

12 yeara---------------13 years---------------14 years---------------15 years--.-..-------~6 years ------------

17 years---------------

‘n

540542526525496416

454490484424;:;

101

%84

:;

88

1!:

;:74

N

1,7471,7291,6821,6461,5941,525

1,6811,6671,6331,5901,5421,502

280262256241231225

272275266235243237

6.56.86.97.27.07.8

:::10.311.111.411.1

4.95.r

H

u

8.8

%:

.:::

.0.1

5.015.575.925.845.015.65

5.916.336.287.137.067.11

3.244.39$.::

2:986.06

6.026.035.396.026.025.53

0,220.200.330.260.280.34

0.340.36C&J

0:380.52

0.360.440.540.190.340.72

0.730.560.360.480.910.64

Percentile

5th 10th 25th 50th 75th 90th 95th

In millimeters

3.03.23.23.43.53.6

::::.:.

:::

3.13.0

;:;3.23.5

3.7

2:?:.;

5:0

3.2

H

:::4.1

4.24.55.15.25.65.4

3.23.23.23.5

:::

4.34.2

:::;.;.

:::4.24.44.54.6

5.1

2:;6.56.96.7

3.5

:::4.24.34.3

5.4

;:;6.16.16.2

4.85.05.25.55.65.8

6.6

;::

R8.8

4.24.7

2::

:::

H8.37.6-

:::

6.87.17.17.6

:::

10.611.512.813.413.713.8

5.15.8

H

%!

9.8

13::11.411.713.7

14.213.814.313.413.215.1

16.818.220.122.521.621.1

1;:;8.27.2

::;

17.318.316.719.221.418.0

19.619.519.819.718.322.4

23,523.323.227.428.226.2

13.215.2;:.:.

1:::

22.523.821.123.423.521.3

NOTE: n = samplesize;N= estimated number of youths in population in thousands; ~- mean;s = standarddeviation;& = standarderror of the mean.

25

Page 31: Skinfold Thickness of Youths12=17Years

!fable10. Suprailtacskinfoldof youtheby race, sex, and age at last birthday:samplesize, es-timated populationsike,mean, standarddeviation,standarderrorof themean, and selectedper-centiles,United States,1966-70

Race, sex, end age

Male

years -------------years -------------years-------------years-------------years -------------years -------------

Female

years -------------years-------------years-------------years-------------years-------------years-------------

NEGRO

Male

years-------------years-------------years-------------years-------------years-------------years-------------

Female

years-------------years-------------years-------------years-------------years-------------years-------------

n

538542527525495415

453490484424441392

1008088845769

::101

;:74

N

1,7441,7291,6861,6461,5921,520

1,6771,6671.6331,5901,5421,499

278262256241231225

272275266235243237

z

12.312.612.613.213.214.4

13.914.915.716.716.216.1

:::9.0

;::9.9

13.613.214.414.315.914.6

8

9.259.639.499.288.759.92

8.208.067.788.517.958.66

6.868.158.225.495.758.27

9.279.097.487.509.208.42

8X

.

0.320.400.480.380.520.64

0.480.410.380.560.510.57

0.901.001.160.590.500.82

1.230.900.73;.;;

0:98

Percentile

5th 10th 25th 50th 75th 90th 95thI I I I I 1

ti millimeters

i:;4.35.15.15.2

4.95.56.36.67.06.2

H

:::4.14.3

4.0

2;

2;5.5

4.54.85.05.65.75.8

5.66.77.47.68.27.4

:.;

4:1

~:;.

5.2

2:!6.46.97.2

5.96.16.57.27.27.6

7.7

1;:$10.510.810.2

4.95.05.2

;:;5.5

;::8.6

;::9.5

9.19.1

Q:::10.311.1

12.313.414.214.915.114.1

6.26.66.86.5

;::

11.310.913.7;;.;

12:8

16.516.215.615.716.818.3

18.719.320.522.120.220.9

1?::

1;::10.810.5

18.717.018.718.221.317.7

;;.;

27:527.625.930.2

25.526.126.829.727.228.6

18.222.415.613.415.419.8

25.827.725.423.830.824.7

31.834.433.634.232.838.3

30.832.031.334.531.933.0

23.533.326.123.419.626.6

32.336.630.229.535.530.4

NOTE: tl. samplesize; I?-estimatednumber of youths in populationin thousends; X. mean;8 = standarddeviation;SE= standarderror of the mean.

26

Page 32: Skinfold Thickness of Youths12=17Years

Table 11. Medial calf skinfold of youths by race, sex, and age at last birthday: sample size, es-timated population size, mean, standard deviation, standard error of the mean, and selected ~er-centiles, United States, 1966-70

Race, sex, and age

WHITE

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

NEGRO

Male

years---------------years---------------yeara---------------years---------------yeara---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

n

537537522522493415

450487478420435389

100

;?83

:;

;:100

;;70

N

L,738L,714L,672L,637L,586L,521

1,667L,6571,610.,576!,521!,487

278251253238231225

253275263233243219

12.512.811.811.210.410.3

15.617.118.219.419.520.5

::;8.27.0

;:;

13.114.015.214.416.414.2

6.906.987.016.575.986.15

6.777.077.017.787.027.72

5.006.795.993.483.385.91

5.427.255.415.123.27k.97

In millimeters

0.280.370.380.340.320.32

0.320.360.400.570.420.53

0.690.820.800.360.430.70

0.840.580.470.561.520.68

5.25.5

::;4.24.2

7.57.7

;:$LO.3LO.2

3.6

::!4.23.53.3

::;

;::7.67.9

2::5.65.5

::;

HLO.5L1.O11.611.7

::;4.34.54.13.6

7.5

;:;

H8.5

;:;7.17.2

::2

:1.3!2.3.3.1.4.5.5.0.5.5

5.25.25.2

H4.9

9.7

;::0.6,1.01.0

11.111.310.210.08.78.8

14.415.917.318.419.220.0

;::

:::

::;

11.411.914.513.814.313.4

16.216.114.913.513.312.5

19.621.723.123.423.824.7

10.610.29.6

;:;8.4

15.817.020.217.12.9.717.9

21.922.820.720.319.619.1

24.726.528.430.228.831.2

16.117.412.410.7L1.73.6.2

20.824.625.324.225.921.2

26.827.525.425.122.623.0

30.431.430.935.631.634.8

19.824.521.214.813.21.8.8

~5.510.6~7*~~7.838.512.7

NOTE: tt = eample size; iV=estimated number of youths in population in thousands; ~=mean;S = standard deviation; Sz= standard error of the mean.

27

Page 33: Skinfold Thickness of Youths12=17Years

Table 12. Skewnesscoefficients (differencebetween the 50thand 95th percentiledividedby thedifferencebetween the 5th and 50th percentile)by skinfoldsite, race, sex, and age at lastbirthdayof youths:United States,1966-70

Sex and age

678910”11121314151617

Male

years-------------------years-------------------years-------------------veals-------------------years------------------years------------------years------------------years------------------years------------------years------------------years------------------years------------------

Female

6 years-------------------7 years-------------------8 years-------------------9 years-------------------10 years------------------11 years------------------12 years------------------13 years------------------14 years------------------15 years------------------16 years------------------17 years------------------

Tricepsskinfold

White

1.672.173.003.002.632.672.982.853.453.883.873.62

1.502.291.802.441.832.002.212.071.811.931.661.71

1.332.002.603.002.003.674.494.974.583.312.325.17

3.503.404.002.502.802.503.274.541.732.293.001.97

Wbscapularskinfold

Jhite

3.002.334.666.676.675.757.307.555.936.076.155.37

3.677.005.006.507.004.335.044.743.904.384.123.93

Vegro

2.503.001.001.504.666.008.125.354.102.691.705.45

2.002.508.004.003.004.334.316.304.213.815.063.34

fidsxillaryskinfold

tiite

2.004.006.00

11.00IS. 006.508.237.747.526.926.097.63

6.004.335.005.004.675.205.60;.;;

4:524.004.16

!Tegro

2.001.332.003.006.009.007.756.233.343.781.88LO.33

3.004.004.673.505.505.504.545.293.074.864.102.97

Suprailiacskinfold

Jhite

4.3(i5.184.814.834.294.59

;.;:

2:172.372.082.39

legro

6.4;7.936.036.843.126.88

-

2.8;;.::

2:143.512.38

Medial calfskinfold

Jhi.te

2.672.822.992.843.143.05

-

2.3;1.921.691.871.641.78

Iegro

3.0;4.994.644.102.234.29

2.903.341.702.173.611.69

28

Page 34: Skinfold Thickness of Youths12=17Years

Table 13. Sample size and correlation coefficients far all p;;~~& pairs of skinfolds of youths by race and sex: United States,

Sate and sex says kTriceps-Calf

Skinfold pair

3ubscap-Ular-nidaxil-lary

.9114

.9100

.9017

.9123

3ubscap-ular-supra-iliac

.8371

fdaxil-lary-supra-iliac

.8523

:ubscap-ular-Calf

!Iidaxil-lary-calf

.7437

Supra-iliac-calf

.7256

.7920

.6S50

.7127

Triceps-mbscap-ular

.S063

rriceps-nidaxil-lary

.8002

rriceps-supra-iliac

——

.7790

+

ITotall

Both sexes----- .7351

.7497

.6890

.7346

.7454

.6956

.7794

.8065

.7224

Male---------------F~ale ........-...-

3,513 .8745

3,172 .8149

.7537

.7004

.7388

.7454

.7005

.7824

.S088

.7s35

.8068

.7951

.7797

.Soos

.S3S0

.7540

.7724

.8691

.8296

.5349

.8709

.S527

.8483

White

Both sexes-----

+Male ---------------

Female-------------

3,020

I .S7012,65o .S185

.7817

.6802

.7883

.s077

.7893

.8179

.7910

.7s79

.7960

.8336

.7545

.8059

.9081

.9054

.9138

.8703

.8289

.8666

.8683

.8524

.8812

.8901

.8627

Negro

Both sexes----- -1-982 .8584

Male---------------

Female-------------

.8135

.7180

.8301

.72S0

.8173

.7816

.8144

.7404

.8464

.7490

.9309

.8904

.8829

.8426

lIncludesdata for “other races,” which are not shc.wnseparately.

29

Page 35: Skinfold Thickness of Youths12=17Years

Table 14. Sample size and correlation coefficients for all possible pairs of skinfolds of youths by race, sax, and age at lastbirthday: United States, 1966-70

Race, sax, and.ase

12

;2151617

12

H151617

121314

::17

121314151617

12131415

U

121314

!:17

T~AL1

~

years-----------years-----------years-----------years-----------years-----------years-----------

ue

years-----------years-----------years-----------years-------7---years-----------years-----------

WIWTE

Male—

years-----------years-----------years-----------years-----------years-----------years-----------

Female

years-----------years-----------years-----------years-----------years-----------years-----------

~(1

Hale—

years-----------years-----------years-----------years-----------years-----------years-----------

Famale

years-----------years-----------years-----------years-----------years-----------years-----------

6366206116095514s5

532579579495529458

535537521522491413

446487478418:;;

99

i?

:;69

82

1%

i;70

:riceps-calf

.S7S6

.8801

.8666

.8874

.8722

.8502

.S286

.8568

.7848:;;;:

.7556

.8752

.8716

.8605

.8885

.8703

.8400

.8437

.8449

.8170

.8189

.7804

.7544

.S725

.9224

.8992

.S089

.8887

.9104

.7164

.9081

.6402

.8201

.8516

.7138

Skinfold pair

‘==E=I==.7651.7847.7598.7941.7288.7939

.7422

.7479

.6218

.6942

.6623

.6142

.7581;;:;;

.7992

.7300

.782S

.7485

.7400

.6370

.7073

.6489

.6348

.S198

.8346

.8757

.6929

.7807

.8979

.7419,s431.5766.6693.8118.6893

.7836

.7757

.7740

.7487

.7510

.7992

.8479

.8152

.7966

.83o6

.7527

.7959

.7492

.7539

.6608

.6965

.6534

.5758

.8394

.8092;;;7;

.7467

.7810

.7568

.7448

.6592

.6895

.6403

.5756

.8419

.8533

.8910

.7147

.7971

.8706

.7039;::::

.7355

.7551

.6689

i’riceps-subscap-ular

.8258

.8511

.S206

.8345

.8232

.8260

.8125

.8161

.7189

.7874

.7723

.7613

.8228;:3;;

.8413

.8241

.8285

.8155

.8152

.7539

.797s

.7636

.7577

.8978

.9124

.S630

.6638

.8346

.8380

.8242

.8604

.5736

.7353

.8566

.s509

‘riceps-!idaxil-lary

.8285

.8177

.8328

.7714

.8271

.8073

.8117

.8112

.7406

.8002

.7555

.7398

.8244

.Slol

.8325

.7709

.8250

.8001

.8256

.8097

.7786

.s053

.7511

.7404

.S984

.9167

.8516

.7106

.8540

.8503

.7296

.8357

.5632

.7s73

.7958

.7S20

rriceps-supra-iliac

.8652

.8780

.8636

.8714

.8126

.8167

.8015

.7993

.7261

.7725

.7387

.6916

.8575

.8761

.8568

.8748

.8106

.8118

.8111

.7869

.7538

.7614

.7375

.6901

.9019

.8868

.8991

.7499

.8120

.8436

.7481

.8464

.5976

.8352

.7662

.7033

hll.l=ap

nidaxillary

.9332

.9100

.9156

.8946

.9017

.9226

.9123

.9074

.8926

.9117

.8849

.8953

.9316

.9054

.9171

.8952;;;;;

.9198

.9025

.8937

.9134

.9008

.8972

.9282

.9673.

.9407

.8747

.8643

.9458

.8581

.9397

.8934

.9096

.8576

.9039

TSubscap- Mfdaxil-ular- lary-stipra- supra-iliac iliac

.8533

.8840

.8794

.8765

.8575

.8701

.8459

.8654

.8198

.8281

.8286

.7978

.8513

.8833

.8797

.s779

.8608

.8814

.8508::;;:

.8259

.8283

.7898

.9019

.9023

.9024

.8723

.8212

.8873

.8270

.9039

.8001

.8549

.8369

.8475

.8857

.8646

.8879

.8322

.8778

.8894

.853o

.8901

.8555

.8558

.8377

.8209

.8807

.8590

.8900

.8312

.8770

.8919

lIncludesdata for ‘)otherraces,” which are not shown separately.

30

Page 36: Skinfold Thickness of Youths12=17Years

Table 15. Rankings of correlation coefficients for all possible pairs of skinfolds of youehs with each race, sex, and age group:United States. 1966-70

Race, sex, and age

TOTAL1

~

12 years-----------13 years-----------14 year8-----------15 years-----------16 years-----------17 years-----------

F-

12 years-----------13 years-----------14 year5-----------15 year9-----------16 year8-----------17 yeax8-----------

WHJH

*

12 year8-----------13 years-----------14 yeara-----------15 ycmx+----------16 years-----------17 years-----------

l?emale

12 yettrs-----------13 years-----------14 years-----------15 years-----------16 years-----------17 years-----------

NEGRO

~

12 yeaz9-----------13 years-----------14 years-----------15 years-----------16 years-----------17 years-----------

Female

12 years-----------13 years-----------14 years-----------15 years -----------16 years -----------17 ycare----r ------

532579579495529458

446487478418636387

Skinfold pair

Criceps-calf

88

;87

777

;6

87

;87

7

;776

48s7

108

286675

3:;IJ=:P -

calf

12

i

i

111132

121312

111232

2

:217

:3152

iidaxil-lary-calf

212123

322323

212133

3

:313

136

:5

4

:233

Supra-iliac-Calf

!riceps-;ubscap-ular

l-iceps-xLdaxil-lary

‘riceps-supra-iliac

;:llyr:p-

]idaxil-Iary

;:;:r:p -

supra-iliac

[idaxil-lary-supra-iliac

969

:9

999999

9

;

;9

;9999

109

:99

;99

109

1Includes data for “other races,,,which ~re not shown separately.

31

Page 37: Skinfold Thickness of Youths12=17Years

Table 16. Triceps skinfold of youths by geographicregion,sex,meted populationsize,mean, standarddeviation,standarderrorStates,1966-70

Geographicregion,sex, and age

1213

ii1617

12

i:151617

121314151617

12131415

i;

$:

14

:;17

121314151617

121314151617

1213

#

17

NORTHEAST

M&eyears---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

MIDWEST

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Femaleyears---------------years---------------years---------------years---------------years---------------years---------------

SOUTH

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Femaleyears---------------years---------------years---------------years---------------years---------------years---------------

WESTU

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

Jn

17316215613712411s

135145139121125106

167163162162152116

15015714810s140132

149149149M;

130

128142161138126117

154152150151130125

13413s138135144113

N

474475466393373364

469430411398390388

569582538571528474

586588530483500536

446459454460481430

430435482474386410

543490489475451496

485493478493.511410

and age at last birthday:sample size, esti-of the mean, and selectedpercentiles,United

Percentile

x s s%5th 10th 25th 50th 75th 90th 95th

In millimeters

11.510.810.O8.9

1:::

13.513.815.115.317.016.9

11.110.710.29.39.28.6

13.714.015.616.417.416.8

10.010.08.38.78.78.7

13.113.214.616.715.815.9

10.010.39.3

::!8.4

11.013.114.014.915.816.1

5.736.416.035.295.486.87

5.956.185.636.976.146.39

6.585.915,685.395.135.41

5.766.226.376.666.906.24

5.315.634.405.344.815.09

6.536.516.726.818.205.58

5.325.355.864.314.234.16

4.545.275.515.806.266.63

0.650.380.600.440.590.57

0.750.540.540.430.520.54

0.540.660.460.600.590.53

0.580.740.520.900.881.06

0.450.470.500.460.380.40

0.460.410.370.781.280.32

0.470.440.640.510.500.52

0.580.570.710.700.411.29

::;4.24.34.54.5

:::7.57.27.67.9

5.1

:::4.54.43.6

Z::

;:;8.59.1

4.44.1

M

u

6.0

‘$;

7:58.1

5.0

::!4.3

‘::;

5.66.5

M7.57.6

6.2

;:;4.85.15.1

;::9.08.3

1::;

H5.2

H4.3

7.8

;:!?

1:::10.5

N4.64.4

t::

6.7

;:?9.49.19.4

5.4

::$

:::4.5

6.47.57.8

H8.8

;::6.46.06.06.4

9.2

lvi11.312.912.6

::;6.56.06.15.4

9.5

1?:211.912.712.2

6.56.35.65.45.65.4

S.6

1:::11.610.612.1

6.6

;:;

:::5.4

8.19.7

1?::11.610.6

10.39.58.47.3

:::

12.512.715.214.217.316.4

9.68.88.87.97.67.2

12.613.114.415.215.916.2

9.08.87.27.37.57.2

:;.:

13:615.913.515.7

8.79.2

;::7.27.1

10.512.312.914.415.415.3

14.113.412.610.712.212.4

16.818.118.318.421.122.0

13.614.113.210.811.910.3

16.917.620.020.621.820.5

13.111.7

1:::11.111.1

17.116.718.320.819.518.7

12.212.811.610.710.211.1

13.515.718.518.419.721.0

20.720.618.216.317.221.7

23.123.423.323.824.824.7

20.619.317.416.516.515.4

23.624.425.926.527.327.2

17.918.5;$;

15:214.4

21.623.222.725.825.423.3

18.519.117.414.914.715.3

17.420.522.623.124.525.4

25.325.122.421.423.028.2

25.325.625.730.229.027.4

24.822.823.422.420.419.7

25.625.528.629.531.329.8

22.322.519.419.819.321.3

25.8;:.:

31:731..326.6

21.521.921.619.41s.316.8

20.725.u23.925.626.427.7

NOTE: ?&- sample size; ~= estimatednumber of youths in populationin thousands;~= mean; S= standarddeviation; SZ= standarderror of the mean.

Page 38: Skinfold Thickness of Youths12=17Years

Table 17. Subscapularskinfoldof youths by geographicregion, sex, and age at last birthday: sample size,esti.metedpopulationsize,mean, standarddeviation,standarderror of the meanUnited States,1966-70

, and selectedpercentiles,

Geographicregion,sex, and age

12

:;151617

NORTHSAST

Male

years--------------years--------------years--------------years--------------years--------------years--------------

Female

years--------------years--------------years--------------years--------------years--------------years--------------

M20WEST

Male

years--------------years--------------years--------------years--------------years--------------years--------------

Female

years--------------yeare--------------years--------------y6ars--------------years--------------years--------------

SOUTH

Maleyears--------------years--------------years--------------years--------------years--------------years--------------

Female.—years--------------vears--------------t: .-

14 ~ears--------------15 years --------------16 years--------------17 years--------------

WEST

Male

12 years ------------13 years --------------14 years --------------15 years --------------16 years--------------17 years--------------

Female

12 years--------------13 years --------------14 years--------------15 years --------------16 years--------------17 years--------------

n

173162156137124117

135145139120125106

167163162162152116

14915714s108140132

149149150163149129

128142161138126118

154152150151131125

134138138135145113

N

474475466393373361

469430411394390388

569582538571528474

583588530483500536

446459456460481427

430435482:;$

412

543490489475453496

485493478493513410

~Percentile

x s Sz5th 10th 25th 50th 75th 90th 95th

8.18.48.38.6

1:::

10.411.211.512.413,313.1

/$:

i::

;:;

10.211.212.212.613.613.6

6.67.8

;::8.48.8

10.310.611.613.212.612.7

7.08.0S.58.2

;::

;:’$10,811.512.713.5

5.716.496.lC5.7?5.125.7t

5.826.756.126.976.486.88

5.806.294.966.064.895.58

6.167.286.797.567.538.12

:.;:

4:785.005.194.55

7.916.S47.948.0S9.147.34

5.455.s76.464.624.474.67

4.426.285.646.036.988.25

0.450.370.600.340.390.45

0.340.490.950.720.770.53

0.320.530.500.420.590.60

0.710.830.700.710.611.27

0.360.330.510,300.420.38

0.600.470.230.561.120.38

0.790.570.670,460.370.66

0.510.630.500.68:.;:

4.14.1

2::5.C5.4

4.85.15.66.0

H

2::4.44.6

2:;

4.5

::!

:::6.2

;::

4:45.05.0

4.44.95.15.85.55.8

:::3.9

:::5.3

4.34.7

H6.26.4

In millimeters

::i4.(5.(5’b:i

5.:5.56.:6.!7.:7.i

::;4.95.1

:;;

H6.76.5

:::

4.24.3

:::5.25.3

4.85.45.66.5

:::

4.04.24.4

H5.8

4.8

2:?6.26.87.0

5.1

;::5.86.17.0

6.27.0

H9.08.4

4.85.25.45.76.36.4

:::8.0

::;8.1

4.7

::;5.55.96.4

6.16.7

:::

;::

4.54.85.3

~:;.

5.76.27.27.77.87.9

6.1:::

b.!

;:;

8.7

:::10.112.211.5

5.76.26.66,.87.37.7

7.9

1:::10.410.811.4

5.66.0

;::7.37.6

7.7

:::10.4

1?::

5.55.96.37.2.

%;

6.9

;:;10.210.911.1

9.18.4

;:;9.:11.1

13.313.513.715.216.616.6

:::

%:

1:::

12.713.214.915.415.918.5

6.87.97.7

;::10.3

11.312.213.615.314.815,1

::;9.2

::;10.5

1::;13.514.414.917.5

14.817.315.115.415.518.(

19.121.219.521.822.322.8

15.913.413.515.415.415.9

21.122.923.224.825.424.4

10.616.413.712.013.214.3

20.820.820.827.324.621.1

11.616.114.712.613.017.4

14.219.217.620.124.623.6

21.621.922.623.921.225.0

23.126.925.729.325.931.6

21.920.216.821.520.319.6

24.327.729.130.231.529.7

18.124.418.316.520.721.0

27.825.426.531.032.328.6

17.721.825.518.017.020.5

17.723.220.924.729.235.1

NOTE: ?l= sample size; ~= estimatednumber of youths in populationin thousands;z = mean; s = standarddeviation; Si= standarderror of the mean.

33

Page 39: Skinfold Thickness of Youths12=17Years

Table 18. Midexillary skinfold of youths by geographic region, sex, and age at last birthday: sample size,estimated population size, mean, standard deviation, standard error of the me-, ,and selected percentiles,United States, 1966-70

Geographicregion,sex, and age

NORTHEAST

Male

12 years---------------13 years ---------------M years ---------------15 years---------------16 years---------------17 years---------------

Female

12 years---------------13 years ---------------14 years---------------151617

121314151617

12

14151617

121314151617

12

~

1617

121314151617

years---------------years---------------years---------------

MIDWEST

~

years---------------years---------------years ---------------years ---------------years ---------------years ---------------

Female

years ---------------years ---------------years ---------------years ---------------years ---------------years ---------------

SOUTH

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

wi

Male

years---------------ykars---------------years---------------years---------------years ---------------years ---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

B

173162156137124117

134145139120125106

167163162162152116

1501.57148108140132

149149150163149130

128142161139126118

154152149151131125

134138138135145113

Iv

474475466393373361

466430411394390388

569582539571529474

586588530483500536

446459456460481430

430435482476386412

543:;;

475453496

485493478493513410

Percentile

x .!? Sz5th 10th 25th 50th 75th 90th 95th

7.17.27.27.47.08.6

9.3

1::?10.411.610.7

6.56.56.57.27.37.4

9.3

1?::11.212.211.6

5.76.26.16.46.77.2

8.59.0

1:::10.010.0

5.96.6

:::6.37.4

7.28.5

1;::10.911.3

;.;;

7:006.275.106.24

5.546.24;.::

::;;.

4.915.395.035.905.395.88

5.90:.::

6:998.087.69

4.454.964.644.844.405.38

7.346.206.667.966.755.55

4.855.105.925.064.195.32

;.::

5:616.216.527.49

0.540.330.880.410.370.59

0.270.350.810.410.530.60

0.210.390.430.460.650.56

0.580.780.570.75:.;:.

0.340.310.450.280.310.52

0.530.430.170.790.580.27

0.620.440.580.560.490.86

0.430.440.520.750.761.34

3.23.23.33.3

u

4.14.14.44.6

::;

3.03.23.33.63.53.4

2:?

2:;

2:;

3.1

::?

$:.

::?

$;

4:5

:::3.13.43.34.0

3.43.54.34.34.94.7

In millimeters

3.43.53.63.6

::;

4.34.8;.;

6:25.9

3.23.4

:::4.03.8

4.54.55.95.45.75.3

3.23.33.33.6

i::

:::4.4

2::5.2

3.13.33.33.8

;::

::;4.84.85.45.3

4.14.2

:::4.64.8

5.26.16.56.37.87.1

::!4.34.54.74.6

5.55.76.86.67.16.5

;:;4.1:.;

4:5

4.75.5;.;

6:26.6

3.6

:::4.44.34.6

4.7:.;

6:36.76,4

5.25.25.2

;:;6.4

7.67.8

::;10.49.2

4.8;.:

5:45.65.7

7.1

;:28.69.69.0

4.64.84.75.15.65.4

6.46.97.68.47.88.8

4.54.94.95.45.45.6

5.97.18.18.59.18.4

7;6

:::7.6

1;:;

12.412.112.412.214.412.9

7.17.07.07.57.6,7.6

12.111.913.613.814,515.5

6;16,,3

::!

7:6

1%;12.115.2u. 911.8

6.2;.:

7:47.07.5

1:::12.212.612.514.9

15.915.214.816.313.517.5

17.118.518.420.518.116.9

15.110.811.413.713.513.5

19.319.822.220.928.322.1

10.413.411.711.011.713.7

23.118.017.824.418.815.8

10.913.314.612.010.615.7

14.014.618.819.420.822.1

21.520.425.122.921.222.7

19.727.121.726.723.121.8

19.818.317.420.519.521.5

24.323.625.027.531.124.9

17.218.714.716.715.618.8

25.523.023.829.625,124.2

18.119.722.217.214.122.6

17.921.221.125.225.427,0

NOTE : n = sample size; N= estimated number of youths in population in thousands; ~- mean; S = standarddeviation; SZ - standard error of the mean.

34

Page 40: Skinfold Thickness of Youths12=17Years

Table 19. Suprailiacskinfoldof youths by geographicregion, sex, and age at last birthday:sample size,estimatedpopulationsize, mean, standarddeviation,standarderror of the mean, and selectedpercentiles,United States,1966-70

Geographicregion,sex, and age

NORTHSASTMale

years---------------years---------------years---------------years---------------years---------------years---------------

Femaleyears---------------years---------------years---------------years---------------years---------------years---------------

MIDWRST

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Female

years---------------years---------------yeare---------------years---------------years---------------years---------------

SOUTHMale

years---------------years---------------years---------------years---------------years-------u-------years---------------

Female

years---------------years---------------years---------------years---------------years---------------years---------------

WEST

Male

years---------------years---------------years---------------years---------------years---------------years---------------

Female

12 years---------------13 years---------------14 years---------------15 years---------------16 years---------------17 years---------------

n

173162156137124117

135145139120125106

167163162162152116

149157148108140131

149149150::;

130

127142161139126118

151152150151131124

134138138135145113

N

47447546639~373361

469430411394390388

569582538571528474

582588530483500532

446459456460479430

426435482;::

412

538490489475453492

485493478493513410

IXs

12.912.812.012.513.115.6

15.015.815.615.716.516.2

12.311.612.413.113.213.4

14.914.616.016.216.715.6

10.511.711.011.312.012.7

13.614.214.916.715.215.3

11.512.712.813.112.214.0

12.014.315.316.616.216.7

9.5310.269.929.199.4310.87

8.678.587.007.247.107.73

9.778.789.129.648.759.13

8.498.117.508.638.408.61

8.099.488.608.798.489.28

8.688.658.669.169.048.69

8.759.479.788.207.359.97

7.157,537.638.327.779.34

I

I Percentile

SF5th 10th 2!ith 50th 75th 90th

0.870.450.980.470.560.61

0.560.521.110.590.520.72

0.490.630.890.620.980.56

0.940.720.791.170.971.16

0.590.850.780.730.930.99

0.660.710.630.941.301.04

0.770.790.730.890.811.71

0.860.690.720.900.861.76

,:::4.L5.(5.>5.(

4.5

:::6.4

:::

H4.3

::;4.7

5,45.1

:::6.55.6

3.6

:::4.34.74.7

4.6

:::6.6

:::

4.1:.;

4:85.15.3

4.55.86.15.7

z::

In millimeters

4.64.94.85.5

R

5.86.87.8

;::8.4

:::5.2

::;5.4

6.26.17.87.4

:::

4.34.44.54.75.45.4

u6.47.67.07.8

4.54.64.6

z::5.8

5.2

2::

;:;7.5

6.1

2:;6.8

;:;

:::10.610.811.410.8

H6.56.8

:::

;::10.5

1::;9.5

5.65.8

::;6.96.8

n8.610.3

1:::

;::6.4

:::7.4

;:;10.310.611.39.7

10.18.7

:::

1::2

13.214.615.314.915.814.5

7.8

;:;

1;:;10.3

13.213.214.814.515.313.8

::2

R

::;

12.012.313.314.012.513.3

;:?

1::210.310.8

10.712.713.315.414.513.6

17.615.314.014.616.120.8

20.319.419.620.520.420.6

16.214.814.916.217.417.8

20.418.520.620.620.721.4

12.614.212.412.415.316.3

18.319.119.322.819.118.6

15.018.115.815.915.416.8

14.618.621.121.820.222.7

27.330.525.527.627.834.4

27.627.725.625.227.226.6

27.724.826.126.825.527.5

25.826.827.330.030.528.4

20.826.226.225.424.028.5

25.427.027.030.327.424.3

26.527.728.725.924.430.5

22.324.526.128.028.330.6

95th

36.035.336.236.633.738.6

31.336.230.431.530.333.3

32.430.731.835.633.229.8

32.530.431.435.034.131.4

28.233.830.430.232.434.3

37.032.435.035.435.531.4

31.934.436.130.428.838.6

27.030.228.931.231.733.2

FTom: n= sample size; iV= estimatednumber of youths in populationin thousands; ~= mean; s= standarddeviation; Si= standarderror of the mean.

35

Page 41: Skinfold Thickness of Youths12=17Years

Table 20. Medial calf skinfoldof youths by geographicregion, sex, and age at last birthday:sample size,estimatedpopulationsize,mean, standarddeviation,standarderror of the mean, and selectedpercentiles,United States.1966-70

Geographicregion, sex, and age

1213

121314151617

121314151617

12

12

17

12

NORTHEAST

we

years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------

Female

years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------

MIDNEST

&

years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------

Female

years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------

fim

~years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------years---------------------------

Female

years---------------------------yeara---------------------------years---------------------------yeara----.--------------~-------years---------------------------years---------------------------

WEST

~

years---------------------------years---------------------------years---------------------------yeara---------------------------years---------------------------years---------------------------

Femaleyears---------------------------years---------------------------years---------------------------years---------------------------years---------------------------years------------------------:--

n

172159155137123118

131145139119124104

166162162162151116

147156144107139131

147147148161149129

124142158138124114

154152147149130124

134136138133143111

N

471466463393370364

457430411392386375

565579538571526474

576584515478497532

441444451454481427

415435471$;;

401

543490481469451493

485487:;:

506402

12.913.012.411.310.811.5

15.917.218.418.320.119.7

12.512.211.911.010.29.0

16.217.018.719.119.720.4

11.211.7

J:;

J:;

15.016.417.219.518.119.0

11.412.211.010.19.39.4

14.015.916.917.818.319.3

6.507.798.557.086.057.30

6.787.056.777.147.067.69

7.496.796.58;.;;

6:07

6.727.607.477.217.637.76

6.396.635.366.355.795.88

7.447.356.888.997.597.29

6.747.006.775.315.425.37

5.916.476.667.316.667.96

0.69D.530.740.630.680.49

0.730.400.980.600.460.36

0.500.780.690.750.540.67

0.510.650.581.050.761.04

0.410.620.390.580.610.68

0.620.690.631.240.911.07

0.580.620.870.570.500.73

0.651.021.150.910.841.24

Percentile

5eh I10th\ 25thI50thl 75th I90th I95th

In millimeters

:::5.5

::;5.1

!::10.710.611.411.3

5.66.0

H

:::

9.9

1:::10.711.510.7

5.2

u5.04.84.7

H

1;::10.411.1

5.35.6

::;4.24.2

8.27.9

1:::11.111.0

8.38.26.87.2

::2

11.211.813.814.1L5.9L5.O

7.67.3

::;6.45.5

11.6L1.912.614.914.415.3

6.8

:::6.36.26.4

10.511.212.413.512.613.4

,-87.08.06.4

H6.0

10.211.612.412.713.813.5

11.711.210.79.7

;:;

14.416.218.317.319.419.0

11.110.810.5

u7.6

14.515.917.618.719.619.7

10.210.38.28,68.18.6

13.815.116.317.717.118.3

-.----

1%!9.5

;:;8.4

13.214.815.617.017.717.8

L6.6L5.4L5.213.413.815.5

L9.722,622.222.025.524.3

15.715.815.313.5L2.610.7

20.621.524.123.622.924.5

14.815.711.512.411.612.1

17.920.522.025.023.523.3

14.615.614.112.611.712.6

17.320.121.222.322.523.7

22.426.422.121.221.121.7

27.426.326.829.229.829.0

23.621.721.520.319.415.0

25.528,229.530.228.632.4

18.821.616.819.318.719.3

24.327.528.031.628.428.2

21.321.820.317.717.816.5

22.123.826.825.928.231.3

26.630.726.827.323.628.8

30.530.230.7:;.;

31:3

27.925.825.625.622.220.8

30.731.733.433.834.734.8

21.624.521.424.022.621.7

33.131.830.541.132.334.3

26.627.824.723.218.921.2

24.4;J:

30:330.135.1

NOTE: ?l= sample size; N= estimatednumber of youths in populationin thousands;~= mean; S- standarddeviation S5= standarderror of the mean.

36

Page 42: Skinfold Thickness of Youths12=17Years

Table 21. Triceps skinfold of youths 12-17 years of age weighing less than 30 kilo-grams, by sex and ageat last birthday: sample size, estimated population sizez mean,standard deviation, standard error of the mean, andStates, 1966-70

selected percentiles, United

N

69155

4412

5;—

I—

n

24

:

124

1;—

Ill Percentile

x s i%

5th 10th 25th 50th 75th 90th 95th

Sex and age

I

Male In millimeters

4.6**

5.3*

5.5

5.3**

5.6*

5.:

6.1**

7.3*

7.i

6.6**

6.8*

6.i

2.41**

1.51*

1.6;

0.61**

0.64*

0.55

2.9**

5.1*

5.6

10.7**

9.0*

9.;

11.6**

9.0*

LO.;

years-------------years-------------years-------------years-------------years-------------years-------------

7.7**

8.4*

8.;

Female

years-------------years-------------years-------------years-------------years-------------years-------------

n= sample size; N=estimated number of youths in population in thousands;.X ~“~%n; S= standard deviation; Sz = standard error of the mean.

Table 22. Triceps skinfold of youths 12-17 years of age weighing 30-34.9 kilograms, bysex and age at last birthda

??’themean, and selected percentiles, United States, 1966-: sample size, estimated population size, mean, standard

deviation, qtandard erroro70

n

::3

431141

N IZsPercentile

S2

5th 10th 25th 50th 75th 90th 95th

Sex and age

In mil meters

2:;

8

15632125

;:?*

8.58.8**

1.942.82

*

3.993.71

**

0.240.72

*

0.391.22

**

-!

5.03.7*

4.53.5**

H*

5.05.2**

R*

6.67.2**

7.47.1*

8.18.6**

8.88.6*

11.310.7

**

1?:;*

13.417.0

**

10.812.8

*

14.117.3

**

years----------years----------years----------years----------years----------years----------

Female

years----------years----------years----------years----------years----------years----------

n== sample size; N=estimated number of youths in population in thousands;~ ~“%in; 8= standard deviation; Si = standard error of the mean.

37

Page 43: Skinfold Thickness of Youths12=17Years

Table 23. Triceps skinfold of youths 12-17 years of age weighing 35-39.9 kilograms,by sex and age atlast birthday: sample size, estimated population size, mean, stand-ard deviation, standard error of the mean, and selected percentiles, ~ited states,1966-70

NPercentile

x s Sz

5th 10th 25th 50th 75th 90th 95th

Sex and age

U

12 years---------13 years---------14 years---------15 years---------16 years---------17 years---------

Female

12 years ---------13 years---------14 years---------15 years ---------16 years ---------17 years---------

In millimeters

8.2

;::**

9.28.78.38.5

9.:

536309106105

329253

::

1:

2.952.86

0.260.340.35

**

2::3.4**

5.4

H4*4

7.;

H5.9**

8.88.68.48.6

8.;

1;.;

7:7**

11.310.710.310.6

9.:

12.412.19.5**

12.611.811.112.8

14.;

14.013.49.9**

14.412.812.516.2

*14.0

2.38**

2.882.812.603.55

2.4;

0.290.380.470.83

1.0:

n= sample size; N== estimated number of youths in population in thousands;~~O%&; s= standard deviation;Sf= standard error of the mean.

Table 24. Triceps skinfold of youths 12-17 years of age weighing 40-44.9 kilograms,by sex and ageatlast birthday: sample size, estimated population size, mean, stand-ard deviation, standard error of the mean1966-70

, and selected percentiles, United States,

N

53633719658243

4283421761749693

Percentile

x s SE

5th 10th 25th 50th 75th 90th 95thn

1331126119

:

118101

%3427

Sex and age

Male In millimeters

12 years-------13 years-------14 years-------15 years-------16 years -------17 years-------

10.08.5

:::6.2*

10.710.210.310.411.611.1

3.643.322.981.781.02

*

3.243.193.603.183.662.34

0.350.260.420.380.54

*

0.280.440.600.470.890.51

:::3.6

%;*

H6.05.96.67.4

2::4.24.04.9*

7.1

:::6.57.67.7

::;5.14.75.4*

8.58.2

U

1::!

10.27.8

:::6.6*

.10.610.210.210.711.411.1

13.110.810.37.67.2*

13.012.412.712.413.212.8

14.813.712.18.67.4*

15.815.314.914.517.314.8

16.415.412.69.07.4*

16.616.3:;.:

19:315.6

Female

12 years-------13 years-------14 years-------15 years-------16 years-------17 years -------

_ NOTE: n= sample size; N=estimated number of youths in population in thousands;X = mean; S-standard deviation; Sx= standard error of the mean.

38

Page 44: Skinfold Thickness of Youths12=17Years

Table 25. Triceps skinfold of youths 12-17 years,of age weighing 45-49.9 kilograms,by sex and age atlast birthdav: sample size, estimated population size. mean. stand-ard d~viat$on,1966-70

standard erro~ of the mean,-and selecte~ ~ercentiles, ~ited-states,

Sex and age

Male

12 years--------13 years--------14 years--------15 years--------16 years--------17 years--------

Female

12 years --------13 years--------14 years--------15 years--------16 years--------17 years--------

n

L!;86

;:12

L06L22L43919076

iv

2993962641858340

372399$;:

293280

IIx s Sz

11.58.97.4

R5.6

12.611.012.012.712.112*2

5.333.443.182.681.851.96

$:;

3:264.104.084.18

0.760.260.320.310.350.59

0.600.300.400.480.560.83

Percentile

5th 10th 25th 50th 75th 90th 95th

In millimeters

6.1

;:;4.44.13.4

7.8

U8.17.47.4

::;5.35.34.54.5

9.7

1:::9.49.49.3

10.38.6

R5.45.6

12.411.011.912.512.211.8

13.811.09.29.26.86.6

15.813.314.715.314.515.4

20.613.212.911.18.28.6

18.215.717.018.718.018.5

22.416.614.413.09.510.0

21.017.217.621.120.920.1

. ~= sample size; ~= estimated number of youths in population in thousands;~-N~’%; s=standard deviation;s%=standard error of the mean.

39

Page 45: Skinfold Thickness of Youths12=17Years

Table 26. Triceps skinfold of youths 12-17 years of age weighing 50-54.9 kilograms,by sex and age at last birthday:samplesize, estimated population size, mean, stand-ard deviation, standard error of the mean , and selected percentiles, United States,1966-70

Sex and age

Male

12 years--------

13 years--------

14 years--------

15 years--------

16 years--------

17 years--------

Female

12 years--------

13 years--------

14 years--------

15 years--------

16 years --------

17 years--------

n

56

88

123

93

56

32

76

97

124

106

136

103

N

161

275

373

278

184

116

273

337

385

388

442

387

I I I

Percentilex s Sg

5th loch 25th 50th 75th 90th 95th

In millimeters

15.0

10.2

8.1

6.8

6.6

5.9

14.3

14.3

13.7

13.7

13.3

15.3

6.06

4.89

4.25

2.36

2.61

2.03

3.99

4.47

4.92

3.62

4.32

4.85

0.92

0.36

0.30

0.25

0.42

0.50

0.61

0.41

0.36

0.44

0.33

0.54

6.4

4.8

4.2

4.3

3.9

3.1

9.3

7.8

7.4

8.4

7.9

7.8

7.4

5.4

4.5

4.6

4.2

4.0

10.2

9.4

8.4

9,5

8.6

9.7

10.3

6.6

5.5

5.5

5.0

4.7

11.5

11.1

10.2

11.4

10.6

12.2

14.8

9.2

7.4

6.6

6.3

5.8

14.2

13.9

12.9

13.7

12.8

15.2

20.2

13.5

9.6

8.3

8.1

7.4

17.0

17.0

16.8

15.9

16.2

18.5

24.3

18.1

14.0

10.4

10.7

8.4

20.8

21.1

20.4

18.9

19.5

21.8

26.0

20.5

15.5

11.2

13.2

10.2

22.5

24.2

22.7

20.6

21.3

24.2

X:”%L;n= sample size;N=estimated number of youths in population in thousands;

S = standard deviation; Sz = standard error of the mean.

40

Page 46: Skinfold Thickness of Youths12=17Years

Table 27. Triceps skinfold of youths 12-17 years of age weighing 55-59.9 kilograms,by sex and age at last birthday sample size, estimated population size, mean, stand-ard deviation, standard error of the mean, and selected percentiles, United States,1966-70

Sex and age

Male

12 years-------

13 years-------

14 years-------

15 years-------

16 years-------

17 years-------

Female

12 years-------

13 years-------

14 years-------

15 years-------

16 years -------

17 years -------

48

69

106

120

108

67

51

81

98

113

105

105

N

139

223

346

396

367

243

182

275

347

406

342

409

x s s%Percentile

5th 10th 25th 50th 75th 90th 95th

In millimeters

16.4

12.2

8.4

7.0

6.6

6.0

17.8

16.2

15.7

15.8

16.3

16.4

6.17

6.16

4.37

3.22

2.39

1.76

5.91

4.66

4.18

4.54

4.92

4.60

0.99

0.76

0.48

0.34

0.19

0.26

0.90

0.54

0.41

0.42

0.52

0.43

6.2

5.2

4.6

4.0

3.8

4.1

9.1

8.8

10.3

9.6

8.2

9.8

7.6

5.6

5.2

4.3

4.5

4.4

10.2

10.6

10.8

10.6

10.2

10.6

12.8

7.5

6.1

5.3

5.4

5.1

13.6

13.5

12.3

13.1

13.2

12.7

16.4

11.1

7.3

6.6

6.4

5.7

18.0

16.3

15.2

15.8

16.5

16.6

22.2

17.5

10.1

8.2

7.8

7.2

23.0

19.5

19.2

18.4

20.3

19.8

24.5

20.5

15.1

10.7

10.2

8.6

25.6

22.7

21.7

22.4

23.3

22.7

26.4

22.5

18.4

12.6

11.3

10.0

28.3

25.6

22.7

24.3

25.2

25.3

= sample size; N=estimated number of youths in population in thousands;X=NO%&;nS= standard deviation; Sf = standard error of the mean.

41

Page 47: Skinfold Thickness of Youths12=17Years

Table 28. Triceps skinfold of youths 12-17 years of age weighiqg 60-64.9 kilograms,by sex and age at last birthday:samplesize, estimated population size, mean, stand-ard deviation, standard error of the mean, and selected percentiles, United States,1966-70

Sex and age

~

12 years -------

13 years-------

14 years-------

15 years-------

16 years-------

17 years-------

Female

12 years-------

13 years-------

14 years-------

15 years---=---

16 years-------17 years-------

n

18

52

72

114

117

108

29

41

61

52

74

74

N

49

169

231

350

378

397

100

133

207

192

263

263

20.3

13.6

10.0

7.9

7.1

6.7

21.7

20,1

18.7

17.8

18.6

17.9

7.90

6.90

4.92

3.24

2.96

2.61

5.29

4.86

4.55

4.86

4.68

5.02

2.30

1.34

0.52

0.30

0.24

0.29

0.87

0.71

0.45

0.86

0.69

0.75

Percentile

5th 10th 25th 50th 75th 90th 95th

7.6

5.2

4.6

4.6

3.6

3.7

12.6

11.8

12.6

12.1

12.3

10.9

In millimeters

9.5

5.5

5.5

5.1

4.2

4.2

15.2

12.9

13.7

12.6

12.8

11.5

.2.7

7.5

7.1

5.6

5.3

5.1

,8.5

,7.1

,5.7

.4.2

4.8

,4.1

22.3

13.3

8.2

7.3

6.6

6.2

22.3

20.7

18.3

17.2

19.6

18.3

27.4

20.2

12.7

10.1

8.7

8.4

25.6

24.2

22.4

20.8

21.9

22.1

29.7

24.5

18.2

12.7

12.1

10.3

26.8

25.5

25.2

25.1

24.7

24.3

33.7

26.6

20.4

15.5

12.9

12.3

29.4

28.1

27.0

25.9

26.5

25.9

NOTE: n =sample size; N=x

estimated number of youths in population in thousands;= mean; S= standard deviation; SiB= standard error of the mean.

42

Page 48: Skinfold Thickness of Youths12=17Years

Table 29. Triceps skinfold of youths 12-17 years of age weighing 65-69.9 kilograms,by sex and age at last birthda~ sample sizeard deviation,

, estimated population size, mean, stand-standard error of the mean, and selected percentiles,

1966-70United States,

Sex and age

Male

12 years -------

13 years -------

14 years -------

15 years -------

16 years -------

17 years -------

Female

12 years -------

13 years -------

14 years -------

15 years -------

16 years -------17 years -------

n

10

23

59

83

96

91

10

27

31

27

36

35

N

30

80

188

269

327

319

34

83

100

102

145

132

24.0

14.5

12.0

9.6

9.1

7.8

24..5

22.2

21.4

19.9

19.5

17.8

Ts s.x

4.38

6.66

5.16

4.54

3.69

3.44

6.63

4.11

4.91

6.18

4.08

4.71

1.84

1.52

0.70

0.52

0.44

0.51

2.69

0.79

1.28

1.12

D.66

1.00

Percentile

5th 10th 25th 50th 75th 9Oth 95th

In millimeters

16.6

6.2

5.8

4.3

5.2

4.3

14.6

16.3

12.3

10.5

13.2

12.2

20.1

7.0

6.4

5.1

5.5

4.7

17.2

17.1

13.4

13.0

13.8

12.6

20.5

10.7

8.2

7.1

6.9

5.7

20.8

20.4

18.0

16.3

16.7

14.1

27.0

12.1

11.4

8.7

7.8

7.1

24.1

22.3

22.4

20.3

20.6

17.9

28.2

19.5

15.5

12.3

11.7

9.8

30.3

25.1

25.4

23.6

22.6

21.2

30.0

26.2

19.2

15.4

14.9

13.1

31.9

28.1

28.1

29.3

25.3

24.6

30.0

27.5

20.8

21.2

16.5

14.7

39.0

28.8

28.3

30.6

26.2

27.6

n=sample size; N= estimated number of youths in population in thousands;~ ~O%&; s =standard deviation; S% = standard error of the mean.

43

Page 49: Skinfold Thickness of Youths12=17Years

Table 30. Triceps skinfold of youths 12-17 years of age weighing 70-74.9 kilograms,by sex and age at last birthday:sample sizeard deviation,

, estimated population size, mean, stand-standard error of the mean, and selected percentiles, United States,

1966-70

Sex and age

Male

12 years -------

13 years -------

14 years -------

15 years-------

16 years -------

17 years -------

Female

12 years -------

13 years -------

14 years -------

15 years -------

16 years -------

17 years -------

n

4

12

26

53

55

68

8

8

25

21

18

16

N

1046

81

164

181

245

28

24

73

92

68

67

1X.s

*

18.4

13.2

11.3

11.0

9.5

24.1

24.9

24.2

21.7

23.8

24.8

*

4.97

5.03

4.84

4.76

4.27

4.22

3.69

5.31

6.16

5.44

5.33

s%

*

1.46

1.06

0.90

0.75

0.44

1.67

1.97

1.24

1.83

1.38

1.07

Percentile

5th 10th 25th 50th 75th 9Oth 95th

In millimeters

*

11.3

6.4

6.1

5.4

4.3

17.4

18.7

17.4

7.4

15.5

17.2

*

11.6

7.2

6.5

6.3

5.0

17.7

21.0

18.8

7.9

17.2

20.0

*

15.3

9.2

8.1

7.6

6.0

22.0

21.7

21.2

19.7

20.1

21.5

*

17.3

13.4

10.2

10.1

9.6

25.3

26.9

23.5

22.8

24.4

23.6

*

22.3

17.6

15.2

13.8

12.0

28.1

28.2

27.8

25.5

28.4

27.7

*

22.9

21.1

17.8

18.0

14.7

30.0

30.0

31.6

27.9

31.3

33.1

*

26.8

22.4

20.8

19.8

16.5

30.0

30.0

32.6

29.6

34.1

39.0

n= sample size; N=estimated number of youths in population in thousands;-#$:;n; S= standard deviation; s% = standard error of the mean.

44

Page 50: Skinfold Thickness of Youths12=17Years

Table 31. Triceps skinfold of youths 12-17 years of age weighing 75-79.9 kilograms,by sex and age at last birthday: sample size, estimated population size, mean, stand-ard deviation, standard error of the mean, and selected percentiles, United States,1966-70

Sex and age

Male

12 years-------

13 years-------

14 years-------

15 years-------

16 years-------

17 years-------

Female

12 years -------

13 years-------

14 years -------

15 years-------

16 years -------

17 years -------

3

12

27

21

39

51

4

6

6

11

12

8

11

30

85

58

127

182

11

21

20

36

35

37

x

*19.2

14.3

15.8

13.0

12.0

*

22.8

25.7

29.1

26.1

27.0

Percentile

C9 $E

5th 10th 25th 50th 75th 90th 95th

In millimeters

*

5.21

6.59

7.71

6.26

4.71

*

4.08

5.03

4.38

4.57

7.15

*

1.91

1.97

1.51

1.32

0.56

*

2.80

2.56

1.58

2.08

4.72

*

12.8

7.4

8.1

5.4

5.6

*

13.4

16.3

21.8

19.5

15.4

*

13.1

8.2

9.0

5.9

7.0

*

13.9

16.6

22.4

20,0

15.8

*

13.6

10.5

10.3

8.1

8.7

*

20.6

23.7

27.4

22.0

22.3

*

19.6

13.8

13.5

11.0

11.8

*

25.1

25.7

29.6

28.0

28.3

*

23.0

16.0

20.5

18.6

14.7

*

25.7

27.7

31.8

29.5

32.5

*

25.7

21.8

28.5

23.1

16.5

*

27.0

33.0

32.9

31.4

36.0

*

29.0

32.1

29.5

23.8

22.2

*

27,0

33.0

38.0

35.0

36.0

n= sample size; N-estimated number of youths in population in thousands;~~”%m; S=stand.rd deviation; L+ = standard error of the mean.

45

Page 51: Skinfold Thickness of Youths12=17Years

Table 32. Triceps skinfold of youths 12-17 years of age weighing 80-84.9 kilograms,by sex and age at last birthdaygsamplesize, estimated population size, mean, stand-ard deviation, standard error of the mean, and selected percentiles,1966-70

United States,

Sex and age

Male

12 years-------

13 years-------

14 years-------

15 years-------

16 years-------

17 years-------

Female

12 years-------

13 years-------

14 years-------

15 years-------

16 years-------

17 years-------

n

3

7

16

24

17

1

8

3

9

10

5

N

8

21

53

69

65

3

25

9

32

38

16

x

*

17.8

17.1

15.3

13.0

*

29.5

*

27.1

31.2

23.6

Percentile

s Sz

5th 10th 25th 50th 75th 90th 95th

In millimeters

*

2.87

4.27

5.56

4.53

*

7.42

*

2,08

7.30

5.98

*

1.22

1.37

1.28

1.28

*

3.19*

0.79

2.66

3.51

*

14.6

11.4

6.4

7.1

*

17.6*

24.4

16.8

17.1

*

14.7

11.8

6.9

7.8

*

20.1

*

24.8

23.6

17.2

*

15.6

15.3

12.2

9.8

*

20.9*

25.6

26.0

17.6

*

18.2

17.2

14.6

12.4

*

28.6*

27.5

31.1

23.4

*

20.5

18.5

20.6

17.4

*

36.2*

28.6

40.0

31.0

*

23.0

26.2

23.1

19.7

*

40.0*

30.6

40.0

31.3

.

*

23.0

28.0

25.0

20.7

*

40.0*

32.0

40.0

31.4

_ NOTE: ?l= sample size; ~= estimated number of youths in population in thousands;X = mean; s= standard deviation; SX =standard error of the mean.

46

Page 52: Skinfold Thickness of Youths12=17Years

Table 33. Triceps skinfold of youths 12-17 years of age weighing 85-89.9 kilograms,by sex and age at last birthday:sample size, estimated population size, mean, stand-ard deviation, standard error of the mean, and selected percentiles,1966-70

United States,

Sex and age

Male

12 years ------

13 years ------

14 years ------

15 years------

16 years ------

17 years ------

Female

12 years ------

13 years ------

14 years ------

15 years ------

16 years ------

17 years ------

n

5

5

13

10

15

2

2

7

7

5

N

13

13

35

32

47

8

5

32

22

17

I I IIll Percentile

x s s%

5th 10th 25th 50th 75th 90th

In millimeters

>’t

*

22.6

14.4

14.9

-k

*

30.9

25.4

30.7

*

>’?

5.61

4.96

5.74

*

*

4.68

10.55

2.49

+;

*

1.68

1.68

1.92

*

*

1.34

$.70

L.64

*

-k

9.5

9.8

7.8

*

*

22.4

9.7

28.3

+C

*

16.3

10.2

8.3

*

*

22.9

9.8

28.5

>’t

*

21.0

10.8

12.3

*

*

28.1

17.4

29.3

*

*

21.9

12.6

14.2

*

*

32.5

25.4

30.3

*

$<

29.0

19.1

21.1

*

*

33.7

35.4

30.9

*

*

29.8

20.7

25.6

*

*

34.9

40.0

35.0

95th

*

*

30.0

26.2

26.0

*

*

39.0

40.0

35.0

~ = sample size; N=estimated number of youths in population in thousands;Z ~“~%m; S= standard deviation; S% = standard error of the mean.

47

Page 53: Skinfold Thickness of Youths12=17Years

Table 34. Triceps skinfold of youths 12-17 years of age weighing 90-99.9 kilograms,by sex and age at last birthday:samplesize, estimated population size, mean, stand-ard deviation, standard error of the mean1966-70

, and selected percentiles, United States,

Sex and age

Male

12 years-------

13 years-------

3.4years-------

15 years-------

16 years-------

17 years-------

Female

12 years-------

13 years-------

14 years-------

15 years-------

16 years-------

17 years-------

n

1

2

7

8

11

19

4

3

2

9

6

N

2

7

26

26

37

78

10

10

6

29

20

I

*

*

27.4

16.7

17.7

20.7

*

*

*

31.8

32.9

*

*

7.25

4.05

4.08

6.93

*

*

*

4.60

7.56

Percentile

SE

5th 10th 25th 50th 75th 90th 95th

In millimeters

*

*

4.64

1.74

1.94

2.93

*

*

*

1.48

4.03

*

*

15.5

10.7

11.7

9.5

*

*

*

26.2

17.6

*

*

16.0

10.8

11.9

11.7

*

*

*

26.5

17.7

*

*

22.5

14.2

13.8

15.7

*

*

*

27.4

31.8

*

*

32.2

17.1

17.5

20.4

*

*

*

32.3

34.2

*

*

37.0

21.4

24.1

32.0

*

*

*

40.4

40.4

*

*

37.0

21.4

24.3

32.0

*

*

*

40.0

40.4

. = sample size; ~=estimated number of youths in population in thousands;X ~“%m;

nS= standard deviation; Sz = standard error of the mean.

48

Page 54: Skinfold Thickness of Youths12=17Years

Table 35. Triceps skinfoldofyouths 12-17 years of age weighing 100 kilograms or more,by sex and age at last birthday sample size, estimated population size, mean, stand-ard deviation, standard error of the mean, and selected percentiles, United States,1966-70

Sex and age

Male

12 years-------13 years-------14 years-------15 years-------16 years-------17 years-------

Female

12 years-------13 years-------14 years-------15 years-------16 years-------17 years-------

nPercentile

z s’ s%

5th 10th 25th 50th 75th 90th 95th

In millimeters

*

21.:**

****

*

6.8;**

****

*

2.6:**

****

*

12.:**

****

*

12.:**

****

*

14.;**

****

*

20.;**

****

*

26.;**

****

*

34.:**

****

*

34.;

*

****

- NOTE: ‘??=sample size;lV=X= mean; S=

estimated number of youths in population in thousands;standard deviation; S: = standard error of the mean.

49

Page 55: Skinfold Thickness of Youths12=17Years

Table 36. Area of arm at triceps, estimated area of muscle and bone,,andestimated areaof fat for males and females 6-17 years of age: United States, 1963-70

I I

Sex and age

Male

6 years------------------------------------------

7 years------------------------------------------

8 years------------------------------------------

9 years------------------------------------------

1011

12

13

1415

16

17

years-----------------------------------------

years------------.----------------------------

years-----------------------------------------

years-----------------------------------------

years-----------------------------------------

years---------------.-------------------------

years-----------------------------------------

Female

6 years------------------------------------------

7 years------------------------------------------

8 years------------------------------------------

9 years------------------------------------------

10

11

12

13

14

15

16

17

years-----------------------------------------

years-----------------------------------------

years-----------------------------------------

years-----------------------------------------

years-----------------------------------------

years-----------------------------------------

years-----------------------------------------

years-----------------------------------------

EstimatedAr~myf area of Estimated

muscle and area of

bonel fats

I I

In square millimeters

2,410

2,586

2,764

2,964

3,135

3,454

4,051

4,547

5,127

5>645

6,000

6,348

2,380

2,565

2,828

3,043

3,332

3,5734,242

4,471

4,775

5,142

5,119

5,231

1,761

1,926

2,052

2,183

2,320

2,552

3,011

3,452

4,089

4,623

4,968

5,233

1,662

1,754

1,901

2,068

2,214

2,436

2,976

3,092

3,207

3,367

3,350

3,331

649

660

712

781

815

902

1,040

1,095

1,038

1,022

1,032

1,115

718

811

927

975

1,118

1,137

1,266

1,379

1,568

1,775

1,769

1,900——

1.

2%%;&n;ed area of fat = area of arm - estimated area of muscle and bone.

50

Page 56: Skinfold Thickness of Youths12=17Years

Table 37. Comparison of skewness ratios between races and sexeslfor skinfolds of chil-dren and youths: United States, 1963-70

Comparative race or sex group

White ratio greater than Negro ratio ------

Negro ratio greater than white ratio ------

White ratio equal to Negro ratio ----------

Male ratio greater than female ratio------

Female ratio greater than male ratio ------

Male ratio equal to female ratio ----------

Triceps

7

16

1

18

6

0

Medialcalf

2

10

0

11

1

0

Skinfold

Sub-scapular

15

8

1

13

11

0

Midax-illary

16

7

1

15

10

0

lWmber of times one ratio was greater than another, comparisons made byeither race or sex.

Supra-iliac

4

8

0

11

1

0

age and

51

Page 57: Skinfold Thickness of Youths12=17Years

APPENDIX I

STATISTICAL NOTES

The Survey Design

The sampling plan of the third cycle of the HealthExamination Survey followed a multistage, stratifiedprobability sample of clusters of households in land-based segments in which a sample of the United Statespopulation (including Alaska and Hawaii) aged 12through 17 years was selected. Excluded were thoseyouths confined to institutions and those residing uponany of the reservation lands set aside for use by Amer-ican IncWms.

The sample design of Cycle III is similar to thatof Cycle H in that it utilizes the same 40 sample areasand the same segments. The decision to incorporatethis feature into Cycle III was not made prior to theselection of the second cycle sample although it is con-sistent with the early concept of a single program for6-17 year olds. The final decision to utilize this iden-tical sampling frame was made during the operationof the second cycle program.

The successive elements for this sample designare primary sampling uni~ census enumeration dis-tric~ segment (a cluster of households); househol~all eligible youths; and finally, sample youth. Everyeligible youth within the defined population has a knownand approximately equal chance for selection into thesample.

The steps of drawing the sample were carried outjointly with the Bureau of the Census; the startingpoints were the 1960 decennial census lists of addressesand the nearly 1,900 primary sampling ‘units (PSU’S)into which the entire United States was divided. EachPSU is a standard metropolitan statistical area (SMSA),a county, or a group of two or three contiguous coun-ties. These PSU’S were grouped into 40 strata so thateach stratum had an average size of about 4.5 millionpersons. This grouping was done in a manner whichmaximized the degree of homogeneity within stratawith regard to the population size of the PSU’S, degreeof urbsrdzation, geographic proximi~, and degree ofindustrialization. The 40 strata were then classifiedinto four broad geographic regions of 10 s~ata eachand then within each region, cross-classified by fourpopulation density classes and by tlie rates of popula-tion change from 1950 to 1960. Using a modified Good-

52

man-Kish controlled-selection technique, one PSU wasdrawn from each of the 40 strata.

The sampling within PSU’S was carried out in sev-eral steps. The first was the selection of census enu-meration districts (ED’s). These ED’s are small well-defined areas of about 250 housing units into which theentire Nation was divided for the 1960 population census.Each ED was assigned a “measure of size” equal to therounded whole number resulting from a “division bynine” of the number of children aged 5-9 in the ED atthe time of the 1960 census. A sample of 20 ED’s inthe sample PSU was selected according to a systematicsampling technique with each ED having a probabilityof selection proportional to the population of children5-9 years at the time of the 1960 census date. From eachED a random selection of one measure of size (seg-ment) was taken.

Minor changes required in the Cycle III design werethat it be supplemented for new construction to a greaterextent than had been necessary in Cycle II and thatreserve segments be added. Although it was the planfor Cycle III to use the Cycle H segments, it was rec-ognized that within several PSUfs, additional reservesegments would be needed to avoid the risk of havingan insufficient number of examinees. This was promptedby the fact that four of the PSU’S in Cycle II had yieldsof less than 165 eligible children and several otherswere marginal in their yield. In addition, there was a3-year interval between Cycle II and Cycle III, so thatit was quite possible for some segments to have beencompletely demolished to make room for highway con-struction or urban redevelopment.

The time available for examinations at a particularlocation or stand, as they have been designated, isnecessarily set far in advance of any preliminary fieldwork at the stand. Therefore, the number of examina-tions that can be performed at a particular location isdependent upon the number of examining days available.At the majority of locations the number of days avail-able, excluding Saturdays, is 17. At the rate of 12 ex-aminations each day, this provides for 204 examinationslots. Examinations are conducted on Saturdays if forsome reason it is necessary. Because of reschedulingfor cancellations or no-shows, the maximum numberof youths that is considered for inclusion in the sample

Page 58: Skinfold Thickness of Youths12=17Years

is 200. When the number of eligible youths exceedsthis number, subsampling is performed to reduce thenumber to manageable limits. This is accomplishedthrough the use of a master list which is a listing ofall eligible youths in order by segment, serial number(household order within segment), and column number(order in the household by age). After the subsamplingrate has been determined, every rsth name on the listis deleted, starting with the yth name, y being a rsn-domly selected number between 1 and n. Youths whoare deleted from the Cycle III sample but who wereexamined in Cycle 11as well as any twin who may havebeen deleted are, if time permits, scheduled for anexamination for inclusion only in the longitudinal studyportion or twin study portion of the survey. Their dataare not included in the report aa part of the regularsample.

Since the strata are roughly equal in populationsize and a nearly equal number of sample youtha wereexamined in each of the sample PSU’s, the sample de-sign is essentially self-weighting with respect to thetarget population; that is, each child 12 through 17 yearsold had about the same probability of being drawn intothe sample.

The adjustment upward for nonresponse is intendedto minimize the impact of nonresponse on final es-timates by imputing to nonrespondents the character-istics of “similar” respondents. Here “similar” re-spondents were judged to be examined youths in asample PSU having the same age (in years) and sexas those not examined in that sample PSU.

The poststratitied ratio adjustment used in thethird cycle achieved most of the gains in precisionwhich would have been attained if the sample had beendr~wn from a population stratified by age, color, andsex and made the final sample estimates of populationagree exactly with independent controls prepared bythe U.S. Bureau of the Census for the noninstitutionalpopulation of the United States aa of March 9, 1968(approximate midsurvey point) by color and sex foreach single year of age 12 through 17. The weight ofevery responding sample child in each of the 24 age,race, and sex classes is adjusted upward or down-ward ao that the weighted total within the class equalsthe independent population control.

A more detailed description of the sampling planand estimation procedures is included in Vitul andHealth Statistics, Series 2, Number 43,19 “Sample De-sign and Estimation Procedures for a National HealthExamination Survey of Children,” and in Series 1,Numbers 1,20 5,3 and 8,4 which describe the planand operation of the first three cycles of the HealthExamination Survey (HES).

Some Notes on Response Rates

As mentioned previously, the sample designs ofthe second and third cycles of the HES were similar.

NOTE The list of references follows the text.

Differences did occur, however, in response rates ofvarious subgroups of these samples and these differ-ences deserve some consideration here.

Most importantly, the number of youths selectedfor examination increased from 7,417 in Cycle II to7,514 in Cycle III. The response rate, that is, the num-ber of youths selected who were actually examined,decreased from 96 percent in Cycle H to 90 percent inCycle III. Of the examined youths of Cycle II, 13.86percent were Negro compared with 14.76 percent ofthose examined in Cycle 111. This difference does notreflect a difference in the percentage of Negro youthsselected for examination, but instead, a smaller de-crease in response rate for Negro youths between thetwo cycles than waa the case for the white youths. Inactuality, 13.8 percent of the sample selected for ex-amination was Negro in Cycle III corresponding to 13.5percent for Cycle II. However, whereas the responserate for white youths dropped from 95.6 percent inCycle H to 89.1 percent in Cycle III, the responserate for Negro youths dropped a far lesser degreefrom 98.4 percent to 96.6 percent. Thus, better relativeresponse from the Negro portion of the sample yieldeda greater percentage of these youths actually examinedduring Cycle 111 than was the case during the previoussample.

Examination of sample sizes in this report clearlyshows that at every age group there were fewer girlsactually examined than there were boys of the sameage. This again is not attributed to differences in num-bers of youths selected in the sampling design, butrather to the following differential response rates be-tween maies and females:

Age Male Fatrde

12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93.5 91.313 . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . 93.2 91.914 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91.7 90.715 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91.6 87.916 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89.8 87.717 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87.6 81.8

Total . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . 91.4 88.7

Note that at each age group the response rate for Imysexceeded that of girls.

A similar analysis of response rates can be done

by age, race, and sex as follows:

WhiteAge

Negro Whitemale male female f252

12 . . . . . . . . . . . . . . . . . . . . . . 92.6 99.0 90.1 98.913 . . . . . . . . . . . . . . . . . . . . . . 92.5 98.8 91.1 96.814 . . . . . . . . . . . . . . . . . . . . . . 91.0 97.8 89.6 98.215 . . . . . . . . . . . . . . . . . . . . . . 90.7 97.7 86.4 98.616 . . . . . . . . . . . . . . . . . . . . . . 89.2 95.0 66.6 93.017 . . . . . . . . . . . . . . . . . . . . . . 86.5 95.8 80.2 91.4

Total . . . . . . . . . . . . . . . . 90.5 97.6 87.4 95.8

The above clearly indicates that for all ages underconsideration in Cycle 111of the HES, the response rate

53

Page 59: Skinfold Thickness of Youths12=17Years

for Negro youths exceeded that of white youths of thesame sex and age.

Reasons for differences in response rates aremany but may range from the incentive to get examinedin order to miss a day of school, to fear of the ex-amination itself, to inhibitions with respect to beingexamined. The worst response rate was recorded forthe oldest girls. i.e.. 17-year-old females.

Parameter and Variance Estimation

As each of the 6,768 sample children has anassigned statistical weight, all estimates of populationparameters presented in HES publications are computedtaking this weight into consideration. Thus ,x, the esti-

mate of a population mean,” p,” is computed as follows:

Z=~~Wi Xi/ZWi, where Xl is the observation or

measurement taken on the i th person and Wi isthe statistical weight assigned to that person.

The HES has an extremely complex sampling plan,and obviously the estimation procedure is, by the verynature of the sample, complex as well. A method is re-quired for estimating the reliability of findings which“reflects both the losses from clustering sample casesat two stages and the gains from stratification, ratioestimation, and poststratification.” 2

The method for estimating variances in the HESis the half-sample replication technique. The methodwas developed at the U.S. Bureau of the Census priorto 1957 and has at times been given limited use in theestimation of the reliability of results from the CurrentPopulation Survey. This half-sample replication tech-nique is particularly well suited to the HES becausethe sample, although complex in design, is relativelysmall (6,768 cases) and is based on but 40 strata. Thisfeature permitted the development of a variance esti-mation computer program which produces tables con-taining desired estimates of aggregates, means, ordistributions, together with a table identical in formatbut which contains the estimated variance of the esti-mated statistics. The computations required by themethod are simple, and the internal storage require-ments are well within the limitation of the IBM 360-50computer system utilized at the National Center forHealth Statistics.

Variance estimates computed for this report werebased on 20 balanced half-sample replications. A halfsample was formed by choosing one sample PSU fromeach of 20 pairs of sample PSU’S. The composition ofthe 20 half samples was determined by an orthogonalplan. To compute the variance of any statistic, thisstatistic is computed for each of the 20 half samples.Using the mean as an example, this is denoted ~.

Then, the weighted mean of the entire, undivided sample

NOTE: The list of references follows the text.

{~) is computed. The variance of the mean is the meansquare deviation of each of the 20 half-sample meansabout the overall mean. Symbolically,

and the standard error of the mean is the square rootof this. In a similar reamer, the standard error ofany statistic may be computed.

A detailed description of this replication processby Philip J. McCarthy, Ph. D., has been published.21

Standards af Reliability and Precision

All means, variances, and percentages appearingin this report met defined standards before they wereconsidered acceptably precise and reliable.

The rule for reporting means and percentiles con-sisted of two basic criteria. The first criterion wasthat a sample size of at least five was required. If thisfirst criterion was met, then the second criterion,that the coefficient of variation [i.e., the standarderror of the mean divided by the mean (sX@)] was tobe less than 25 percent, must have been demonstrated.Thus, if either the sample size was too smalI, or thevariation with respect to the mean was too large, theestimate was considered neither precise nor reliableenough to meet the standards established for publi-cation.

To illustrate these criteria, in table 21 all valuesof the distribution of skinfolds for 14-year-old malesweighing less than 30 kilograms were replaced by as-terisks (*) since there were less than five people ofthat age, sex, and weight. Furthermore, in that sametable, although there were five 13-year-old boys, thedistributions of skinfold values are also replaced byasterisks because the standard error with respect tothe mean exceeded the criteria previously stated.

Hypothesis Testing

Although this report is primarily descriptive, itis often desirable to make statistical comparisonsbetween two groups such as males and females or1Z-year-olds and 13-year-olds. Classically, if a statis-tician wishes to test the difference between two m&ms(or, put differently, to test whether two samples couldhave been drawn from the same population), he coulddo so by setting up a normal deviate in which he wouldutilize the means and standard errors of the means ascomputed from the samples. The statistic

is then compared to a table of normal deviates to de-termine whether or not there is, in fact, a difference

54

Page 60: Skinfold Thickness of Youths12=17Years

between the two groups. (Note that the above makes theassumption that the two groups are independent andthat S # + S 3.)

While the technique may appeal to many, in theanalyses of this report this technique is not used fortwo basic reasons:

(1)

(2)

Use of the z statistic makes necessary theassumption of normality. As is clearly shownby the percentile distributions of the variablesconsidered in this repo~ t, this assumption isbadly violated.

Because of the many breakdowns of the HESsample, innumerable tests of this nature couldbe performed and, with each new test, theprobability of rejecting a hypothesis incorrectlymay be .05, but if 10 such tests are performed,the probability of making at least one mistakesomewhere in those 10 tests is somethingcloser to .50.

It was therefore decided to place the greatest em-phasis on a relationship remaining consistent over bothsexes (or races) and all ages under consideration. Inother words, to say that “girls have median tricepsskinfolds greater than boys for all ages between 12 and17 years” has far more meaning and interpretabilitythan to say “the mean triceps skinfold for 12-year-oldgirls is significantly greater than the correspondingmean for 12-year-old boys, and the mean . . . . for 13-year-old girls is significantly greater than tbe meanfor 13-year-old boys, 14-year-old girls, etc.,” as de-termined by a normal deviate. In these analyses, con-sistency rather than a statement about a succession ofindividual probability levels is the factor consideredmost important in demonstrating a relationship.

Analysis of Correlations Among Skinfolds

For each of the 6,768 children in the sample fiveskinfolds were recorded. The correlation coefficientswere computed for each of the 10 possible pairs ofthese five skinfold measurements in the following man-ner:

r= I 2 w, Zw, X,Y, - Zw, X,(zw,y)

d[zw, Zw, Xf-(zw, X,) ’lrxviMi%+’i)’l

where WI is the weight assigned to the i ‘h individualand X and Y are the two skinfold measurements beingcorrelated.

Three correlation coefficients were computed foreach of the 24 age-sex-race categories. The resultsare presented in table 14. As described in the text, itwas decided to rank, witbin each age-sex-race group,the five correlation coefficients under consideration.The distribution of these ranks is shown in table 15.

Imputation

The necessity of arriving at a workable imputationscheme for Cycle 111of the HES was dictated by thefact that each individual carries a separate and uniquestatistical weight, i.e., the number of individuals in theUnited States population he is said to represent. Thedecision to drop from the sample such an individualdue to missing or erroneous values on some numberof variables would not be satisfactory unless the sta-tistical weight was somehow redistributed. The extentof bias introduced in this manner would depend uponthe scheme chosen for the redistribution of the indi-vidual’s statistical weight and would carry along withit the major disadvantage of having unweighed samplesizes differ from variable to variable (thus making cor-relation procedures more complicated), while, ofcourse, the weighted sample sizes would remain con-stant.

A regression method of imputation which was se-lected for the analysis of HES body measurements wasdesirable and possible for several reasons. First, thenumber of problem cases was small enough so as notto be unwieldy. Second, the various body measurementscollected on an individual are highly correlated and,as such, one would like the imputed value to be harmo-nious with the other valid measures for that individual.To simply impute a group mean or a randomly selectedvalue to an atypical individual in place of either a non-existent or an existing but obviously incorrect meas-urement while ignoring the other valid information onthat same individual would be undesirable.

Third, the bias introduced by a regression schemewould clearly be less than would arise if individualswith missing or questionable bits of information wereexcluded from the sample and their statistical weightsredistributed. Fourth, this system has the advantageof holding both the weighted and unweighed samplesizes constant from variable to variable thus facil-itating any correlations or cross-tabulations desired.Thus, an elaborate regression scheme was utilizedto impute body measurements of the third cycle of theHES.

The procedure was as follows: From the total6,768 subjects on whom some body measurements wereperformed, 26 subjects for whom there was one or

more missing value were temporarily dropped andfour files were created from the remaining 6,742 sub-jects. The files were white males, Negro males, whitefemales, and Negro females. It was from these sub-jects that the prediction equations were finally de-veloped.

In a typical case, a subject (for example, a 12-year-old Negro male) might have a body weight recordedwhich is so low as to raise the question of whether therewas an error somewhere in the data preparation proc-ess. However, despite this extremely low value, hisrecord would be otherwise complete. Since all the other

55

Page 61: Skinfold Thickness of Youths12=17Years

variables are recorded for this individual, an estimatefor body weight is derived based on all the other in-formation available and it is possible to conclude thatthe recorded measurement is possible considering theyouth’s other dimensions or that the recorded valueis an obvious clerical error and should be changed.Thus, the file with the Negro males who all have com-plete records is tapped and a stepwise regression iscalculated, with body weight the dependent variable.All the remaining variables are eligible for inclusioninto the equation with the following restrictions:

(1)

(2)

Age must be the first variable added into the equa-tion, irrespective of the correlation between ageand the dependent variable.

So long as adding a new variable contributed atleast .005 (%percent) to the coefficient of multipledetermination (R2), it was included. If the con-tribution was less than that, the equation was fro-zen with all the variables which did add at leastthat much to R2. (No equation included more thaneight independent variables.)

The resulting equation may be of the form

Y= a+~lX~+~zXz+f13Xa+...6k xk

where Y is the predicted sitting height, a, @l,132,B3,etc.are the coefficients generated by the regression, andXl, X2, X3, etc. are the independent variables. By in-serting the recorded values for this subject of X1,X2,X3up to xk ( k being the number of variables contributingsignificantly to R 2,kS8) into the equation, a predictionis arrived at for body weight. A value imputed in thismanner is superior to other possible methods sinceall the relevant information is utilized and allows anextremely large or small person to be assigned a sim-ilarly large or small imputed value.

In actusli~ there were only 19 youths in Cycle 111of the HES whose values for any of the five akinfoldswere either missing or highly questionable on the orig-inal data tape.

NOTE: The list of references follows the text.

To determine whether a skinfold measurement was“questionable,” extremes of the distributions of eachvariable were examined case by case. (Although useful,this procedure allows some highly deviant values to goundetected. For example, hidden in the distribution oftriceps skinfolds maybe an extremely thin individual whohad a mispunched skinfoldfsr too large for his thin buildand MS other four skinfold measurements but never-theless within normal bounds for the entire distributionof all triceps skinfolds from the entire HES sample.)But the magnitude of the problem of bad or missingskinfold data in the HES is very small and oversightssuch as this will not have an appreciable collectiveeffect.

A problem much more prevalent in these skin-fold data than those described above is that of “tightskin.” In some circumstances, the skin is so tightlybound to the underlying tissue that it cannot be pickedup into a double fold by the technician. Code 00.0 wasused by the technician to indicate that he was unableto read the skinfold measurement rather than imply-ing that the skinfold existed but was so small that itmeasured zero.

There were a total of 83 subjects on whom tightskin was recorded for at least one of the skinfolds;and, as can be seen, this most often occurred in themedial calf fold. The frequencies of occurrence oftight skin were as follows:

Of course.

Triceps: 5Subscapular: 6Midsxillary: 4

Suprailiac: 10Medial calf: 67

several vouths had “tight skin” re-corded on more than one of the skinfolds.

A complete description of the problems, the al-ternatives, and the selected procedure for use in im-putation of all the other HES body measurements canbe found in a separate document.22 In addition, a com -plete log wasCycle III datarequest.

kept of all changes made on the originaland these may be made available upon

ooo —

56

Page 62: Skinfold Thickness of Youths12=17Years

APPENDIX II

DEMOGRAPHIC VARIABLES

Regional and demographic characteristics by whichthe population has been classified for this report aredefined as follows.

Age and Sex

Population was classified into 12 age-sex groups—the six ages 12-17 years by sex. Birth certificateswere obtainable for verification of age for 92 percentof the youths. Age stated by the parents was acceptedas the true age for the other 8 percent. Age is ex-pressed as years attained at last birthday.

Race

Skinfolcls were reported by race for white andNegro youths. Youths of other races were not sampledsufficiently for comparison purposes; these youthsrepresented only 0.32 percent of the sample.

Region

Regional data are presented for four regions ofthe continental United States.

Re@”on States Included

Northeast ------- Maine, Vermont, New Hampshire,Massachusetts, Rhode Island,Connecticut, New York,Pennsylvania, New Jersey

Midwest -------- Minnesota, Wisconsin, Michigan,Iowa, Missouri, Illinois, Indiana,Ohio

South ----------- Delaware, Maryland, Virginia,District of Columbia, West Virginia,Kentucky, Temessee,North Carolina, South Carolina,Georgia, Florida, Alabama,Mississippi, Arkansas, Louisiana

West ----------- Washington, Oregon, Idaho, Montana,North Dakota, South Dakota,Wyoming, Nebraska, Kansas,Colorado, Utah, Nevada, California,Arizona, New Mexico, Texas,Oklahoma

57

Page 63: Skinfold Thickness of Youths12=17Years

TECHNIQUES

Introduction

APPENDIX

OF MEASUREMENT

Ill

AND QUALITY CONTROL

The major methodological concerns involved in themeasurement of skinfold thickness are:

In normal, healthy, and well-nourished individuals, ● The calipws utilized. There are a number of cali-as much as 25 percent of the total mass of the bodycan consist of fat cells in quantities large enough toform a definite adipoae tissue. Although a significantproportion of this fat is located internally, oftensurrounding organs such as the kidney, more thanhalf of it is found subcutaneously where it “blankets”the individual. In a number of regions of the tidy theadipose layer may be ,! liftedll with the fingCXS, i.e.,

pulled away from underlying tissues, to form a skin-fold. The skinfold therefore consists of a double layerof subcutaneous &t and $kin whose thickness may bemeasured with appropriate equipment and by exercisingreasonable care (figure I).

pers now available which give comparable results.Figure II illustrates the Lange caliper, now mamtfac-tured by Cambridge (Maryland) Scientific Industries,Inc., and used in the Health Examination Survey. Aswith all acceptable calipers, it is spring-loaded to theclosed position and compresses the fold with a constantpressure of 10 grams/mm.2 throughout its range ofopenings. The calipers are readily calibrated using astandard aluminum step wedge with widths in incre-ments of 10 mm. If the needle indicator strays evenslightly from the exact mark, it can be realigned veryeasily. Extensive data available at Cambridge ScientificIndustries demonstrate that the spring loading is vir-

Bone

Figure 1. Diagram of the technique for measuring a skinfold, a double layer ofsubcutaneous fat and skin. In this case, the triceps skinfold is being measuredwith the Lange caliper. (Drawing courtesy of Muriel Kirkpatrick,Dept. of An-thropology,Temple University)

58

Page 64: Skinfold Thickness of Youths12=17Years

Figure I 1. Lange cal i per .

tually constant and that occasional slight indicator segment of the skin must be “pinched” in order tofluctuation is the only drift in the instrument; when the form a fold than when the adipose tissue is poorlyneedle is realigned, the measurement becomes precise developed, aa it is on the dorsum of the hand. Foragain. a given site the width of the skin should be mini.

● The technique utilized. The most comprehensive real, still yielding a well defined fold.description of the actual technique is given by Bro~ek cas follows (see figure I):

The “skin” ahouldbe lifted bygrasping firmly thefold between the thumb and the forefinger. A firmgrip, not exceeding the pain threshold, eliminatesor at least substantially reduces the variationsin the apparent thickness of skinfold that wouldresult from wide differences in the pulling forceof the fingers.

The width of the skin that is enclosed between thefingers is an important factor. It cannot be stand-ardized, in its absolute size, for all the sites ofthe body. With a thick subcutaneous layer a wider

c From BroZek, J., The measurement of body composition, in M. F. A.Mont agu, e d., .4 n Introduction to Physical Anthropology, ed. 3,1960, pp. 637-686. Courtesy of Charles C. Thomas, Publisher,

Springfield, Illinois.

The depth of the skinfold at which the calipers areplaced on the fold also requires comment. The twosides of the fold are not likely to be parallel, whenthe skin is lifted by one hand, being narrowernear the crest and larger toward the base. Whenthe calipers are placed at the base, the resultingmeasurement is too large. Here, again, the correctdistance J70m the crest is defined as the minimaldistance from the cvest at which a true fold, withsurfaces approximately pwallel to each other andto the contact surfacesof the calipevs, is obtainedupon the application of the calipers to the skin.

Some caliper models only approximate but do notactually achieve the parallelism of the contactsurfaces. However, such parallelism is a desirablefeature of the calipers. In very obese individuals atsome sites no true skinfolds, as defined above,

59

Page 65: Skinfold Thickness of Youths12=17Years

can be obtained. The measurements are still use-ful as indicators of fatness but the “skinfold”measurements are then larger than a double valueof skin plus the subcutaneous layer, taking intoaccount the compression of the tissues by thecalipers. It is recommended to lift the skinfold ata distance of about 1 cm. from the site at which thecalipers are to be placed and the skinfold measured.

● The sites selected for measurement. The thickness

of the subcutaneous tissue may be measured at anynumber of sites, and the choice of a site is dictatedlargely by the problem under investigation. At the sametime, certain sites have become more or less stand-ardized as locations which are readily accessible,which may be more accurately measured, which have alayer of fat of relatively uniform thickness, and whichserve aa a reasonable sample of all the subcutaneousfat of the body. For Cycle 111five sites were selected(1) ts+ce@, over the triceps muscle halfway betweenthe elbow and the acromiaI process of the scapula, withthe skinfold parallel to the longitudinal axis of theupper arm; (2) subscapukw, i cm. below the inferiorangle of the scapula in line with the natural cleavagelines of the skin; (3) mkiiwilkvy, in the midaxillaryline, but with the fold perpendicular to it, midwaybetween the nipple and the umbilicus; (4) sups’ailiac,just above the iliac crest in the midaxillary line, withthe fold perpendicular to i~ and (5) medtizl calf, on themedial aspect of the Iegnear its greatest circumference,the fold running parallel to the long axis of the leg.Diagrams of the five sites are shown in figure III.

Figure Ill. The five skinfold sites: (I) subscapular, (2)triceps, (3) midaxiilary, (+) suprailiac, (5) medial calf(leftcalf skinfoldshownfor convenience).

HES Measuring Technique

Trained observers measured all skinfolds to thenearest 0.5 mm. The values were read aloud to a re-corder, also a trained measurer, who repeated aloudeach number back to the observer as it was recordedin the proper space on the record form. This repetitionserved both as a doublecheck to the measuring tech-nician and to reduce the recorder’s errors. The meas-urement was repeated, and if it did not coincide withthe first one, a third one was taken.

All skinfolds and body measurements were per-formed in a regular sequence to minimize the numberof position changes the child was required to make.The sequence is illustrated on the measurement re-cording form (figure IV).

All of the individuals performing body measure-ments in the HES were experienced X-ray technicianswho had been trained in anatomy and the identificationof specific body landmarks. In addition, X-ray tech-nicians tend to work well with people and are skilledin giving the examinee verbal orders along with thenecessary handling to achieve proper positioning.

Each technician received more than a month offntensive training before being considered profici ent inmaking body measurements. In this training, he be-came skilled with the equipment, the precise locationsof the body at which the measurements were to be taken.and the technique of measurement itself. The majorsources of measurement error by the trainee wereimproper positioning of subject’s body, improper se-lection of specific body landmarks, and improper tech-nique in applying the calipers. Incorrect reading fromthe instrument (usually transposition of numbers) alsooccurred. The measurements of each technician werecarefully compared with those of the other three andwith the measurements of the two supervisors (Dr.Peter V. V. Hamill, the medical advisor, and Dr.Francis E. Johnston, the anthropologic consultant) be-fore they were officially accepted as recordable data.

Broadly conceived, training and quality controlhave two major goals-(1) to substantially reduce thevariability introduced by errors of measurement and(2) to assess the magnitude of the remaining residualerror. The achievement of the first goal requires notonly suitable initial training but also a persistent on-going system of quality control. Achieving the secondrequires the construction of experiments designed toquantify specific components of the error of measure-ment.

Training and quality control for taking body meas-urements consisted of six identifiable procedures, someemphasizing the training component and some theassessment of quality.

(1) Careful training of the examiners.

(2) Periodic direct observation by the medical advisorand the anthropologic consultant as measurements

60

Page 66: Skinfold Thickness of Youths12=17Years

(3)

HEUTH SumAuon SIWEV-111

BODY MEASUREMENTSMttmmensmts is cm. wdsss otherwisespecz~ed.

—l—

zw,—

8-1

12-1

16-1

20-2

n-a

m-c

324!

WOYCUD

SEATEDMEA2.

W.

CERVICAU HEIGtil ———@—

A— HEIGHI ———*—

SADIAL HEIGHT —*—

S7WON HEIGHl ———0—

IUAC CSEST HEIGHT ——e—

1~ HEIGW ———*—

ITBIAL HEIGHT —— — ● —

sum● —

SITTING HEIGHT (Em+) —— —*—

THIGH CLEMAKE ——*—

S?HYRDN HEIGHT — — ● —

roar SRsADm ——o—

I❑ Lmwr

1d TOE REunvEm ad TOS ❑ Shorlw

❑ S’arrw

ANKLE 5SEADTH ——*—

KNEE 5READTH — —*—

ELBOW MEADTH — —*—

WRIST BREAOTH ——*—

2rd (I*) FINGER

I

•l kr

RELATIVE TO ❑ Sk.li.r4* (Nw) FINGER

❑ s...MEDIAL W

SKINFOLO (mm.) ——*—

5TANDINGHEIGHT —— —0—

EMENTS NOT DONE OR SIDE VARIED->

8-10 SIZYGOAWTCMEADTH — —*—

11-13 BIGONIAL 5READTH —0—

14-16 nmw.usow mEArmH—*—

17-19 SEAT 5READTH——e—

~20-2s FOOT LENGTH —m————

I 5TAS4DING

m-as slAcRoMIAlSREAOTH —*—

26-21 SJCRISTALmEAoTH — —*—

!29-31 BITROG+ANTERK 5SEADTH ——0—I

32-34 UWER ABM GIRTH ——*—

3547 FO- GIRTH——*—

SS-61 CHEST GIFTH —— —*—

42-65 WAIST GIRTH——*—

46A9 HIP Glm—0—

50-52 CALF GIKIH——*—

5G-55 TRICEPS SKINFOLD (mm.)—*—

56-58 SUSSCAFWAR SKINFOID (...) — —0—

59-61 LATERALCHEST SKINFOI.D (mm.)

——*—

62-64 SUPDA-IUAC SKINFOLD (mm.)—*—

65-6S WEIGHT (lb.)—0—

I

ijj wbicb and give reason:

~!:a SAJAME NO. ,,.5,

Figure IV. Bodymeasurement record ing form.

were being taken with correction of errors whennecessary.

Practice and retraining during dry runs. The firstday at each location (that is, approximately one daya month) was devoted to dry runs, during which allequipment was retested and recalibrated and reg-ular practice procedures were carried out. Each (4)technician and either a supervising technician orthe supervisors measured one or more people

several times. Discrepancies in measurementswere discussed and any steps necessary to im-prove the techniques were taken. Although thesewere primarily trsining sessions, they afforded anongoing informal assessment as to the quality ofthe data.

Approximately every 6 to 9 months an intensiveevaluation of measurement technique was con-ducted by the supervisors. These sessions lasted

61

Page 67: Skinfold Thickness of Youths12=17Years

(5)

(6)

2 days and involved the measurement, each time,of two boys. One boy was quite fat and the otherwas linear in physique. On the first day both boyswere measured by each of the four technicianswith the supervisors acting as recorders. Thefollowing day the procedure was repeated, thusgiving both inter-observer and intra-observer com-parison of sets of measurements. It was only atthis time that the technicians were allowed to seethe previous measurements andto compare theirswith their ownandwith the other three sets. Majordiscrepancies were noted andattempts were madeto establish the sources of differences and toeliminate them. In addition, such matters as under-lying principles of growth and development and thesignificance of the survey were discussed.

These sessions were intended to include as-sessment of errors due to technician differences,to differences in physiques of subjects, to the siteof the skinfold, as well as to interactions amongthese sources of error. For a variety of reasons,e.g., number of subjects and a greater number oftechnicians in Cycle 111than originally specified inthe model for the analysis of variance, the assess-ment was ultimately abandoned.

A daily instrument chkck was performed on thecalipers using the step wedge as described in thesection “Calipers Utilized,” earlier in this appen-dix.

Several additional calipers were on hand bothto enable the staff to periodically return the instru-ments to the factory for cleaning and for double-checking their precision and to insure against theloss of data in the event of instrument loss ordamage.

The analysis of a set of 301 replicate examinations,taken during 30 stands over the 4 years of CycleIII, provided an estimate of tbe magnitude ofmeasuring error. These data are the subject of thedetailed analysis in the following pages and arejudged to provide a fair estimate of the actualresidual variable measurement error as it oc-

curred during Cycle HI measurements of skinfolds.

Surveillance and Evaluation of Residual

Measurement Process Error

The three following sections are extracted from arecent publication, Quality Control in a National Health.&umination Survey. 22This unusually lucid and well-organized report on quality control was written byWesley Schaible, the quality control officer of tbe HES.Material within brackets has been added to focus thediscussion on skinfold measurements.

NOTE: The list of references follows the (ext.

Monitoring Sysfems

Despite effo~ ts to reduce measurement errors,residual errors of a magnitude large enough to war-rant concern occur with some regularity [in any anthro-pometric survey]. There is, therefore, a real andurgent need to have a system whereby these residualerrors can be monitored. The concept of quality con-trol is based on the desire to obtain end products of acertain quality. Thus, one of the main purposes of amonitoring system is to indicate whether the measure-ments produced by a certain measurement processattain the desired quality. A second purpose is to makepossible quantitative summary descriptions of residual

measurement errors to aid in the interpretation ofsurvey data.

The most extensive system of monitoring used inthe HES in Cycle 111was the collection and evaluationof replicate data. Replicate measurements are usefulfor a variety of purposes—for example, as a means ofincreasing precision of estimates of individual measure-ments, as a training technique, and as a monitoringsystem which includes the objective of final evaluationof measurement errors. These objectives are not in-compatible, and replicate data collected primarily forone of these objectives often indirectly, if not directly,accomplish one or both of the remaining two. For thisreason replicate data are most often collected with acombination of these objectives in mind. The singlemost important source of replicate data in Cycle 111wasthe replicate examinations, in which approximately 5percent of the regular examinees were returned to theexamination center for a second complete examinationexcept for drawing blood and taking X-rays.

Biases ond Controls in Replicate Measurements

A major source of uncertainty in estimates derivedfrom replicate measurements is in the inability to makethe replicate measurement under precisely the sameconditions and in the same manner as for the originalmeasurement. This uncertainty is difficult to evaluateand most attempts to do so are restricted to subjectivestatements concerning the direction and/or size of thebias and the need for concern in the analysis of data.

Several policies regarding Cvcle HI replicate

examinations were specific in tile attempt to obtainmeasurements taken under the same conditions and inthe same manner. Replicate examinations were notconducted during a specific time, but whenever pos-sible were interspersed among the regular examina-tions. An original examination was given priority overa replicate examination in that none was scheduled ifit occupied time needed for a regular examination. Inpractice there was often space to interject replicateexaminations il the schedule without interfering withregular examinations. However, this priority plus thefact that replicates were drawn from those examined

62

Page 68: Skinfold Thickness of Youths12=17Years

increased the likelihood that a replicate examinationwould be scheduled toward the end of the examinationperiod. Nevertheless, the attempt to space the replicateexaminations in the schedule was a valuable policyin that the interspacing of replicate and original exam-inations created an atmosphere more conducive to thereplicate examination’s being conducted in essentiallythe same manner aa the original.

The examiners had been informed of the purposeand importance of the reexaminations. It was emphasizedthat they should not vary their procedures on a repli-cate examination or in any way try to collect “better”data than they normally would. Thereafter, the conductof a replicate examination was not given any greateremphasis than any other instruction since overem-phasizing “sameness” might have created more biasthan it should have eliminated.

At the time of the ori~”nal examination neithev theobscwvey nov tile examinee hnew whether or not thee%minee would be retuvnedfora replicate examination.During the replicate examination, observers were notspecifically informed that an examinee was a replicatealthough no attempt was made to conceal this fact sincein an examination as lengthy as that given in HES theexaminee would undoubtedly be remembered by several,if not all, examiners. Even though an examinee mightbe remembered, it was extremely unlikely that anexaminer would remember a specific measurementafter a time lapse of 2 or 3 weeks. Some bias might beintroduced by the examiner’s knowledge of the replicatestatus of an examinee, but generally this bias wouldseem quite small when compared to the measurementerror and in some cases to the biases associated withthe knowledge and familiarity gained by the examineeduring the original examination. Examinee bias can beimportant, especially in measurements for which aresponse is elicited or when the true value of themeasurement has changed because of a time lapse.Since the time lapse was usually 2 or 3 weeks, someappreciable changes might occur in certain measure-ments such as weight. However, for most of the datacollected the actual change [over time] can only be verysmall, so this effect may usually be neglected. [Forexample, the examinee’s previous experience is muchmore likely to affect, to some extent, the true repli -cability of the psychological tests and those physiologictests requiring high levels of subject participation suchas the treadmill and spirometry; but on those proceduresin which the subject is passive, such as EKG and skin-fold measurements, with very little learning involved,the effect of the previous experience is almost zero inCycle III.]

Replicate data were obtained on approximately 70percent of those selected for such examinations. Oneexplanation for this low rate is that the persuasion andfollowup efforts were not as intensive as for regularexaminees. This is a partial result of giving priorityto regular examinees if interviewer or examination

time was limited. There also seems to be an increasedobjection to returning for a second examination, asdemonstrated in the most frequent reasons for refusal:“One time is enough” and “I can’t miss school again.”

Selection of Replicbte Examinees

The selection of Cycle 111examinees for replicateexaminations was random witbin certain restrictionsimposed by practical considerations. One of the re-strictions was that replicate were selected only fromthose examined during the first week and a half of the

approximately 3?4 weeks of examinations at any onelocation. This time period was chosen to faciiitaie theinterspersing of replicate examinations with originalsin the examining schedule without interfering with thetime allotted for original examinations and withoutscheduling additional time to accommodate replicates.In a voluntary survey it is obviously impossible tofollow a statistically random process in schedulingsubjects, so those scheduled during the first week anda half are not, in the strict sense, a random sample ofall those scheduled, though they may be randomlydistributed for those features which are significant.Evidence that replicates might be considered “repre-sentative” is found in the fact that youths of certainages, locations, incomes, etc., are not routinely morelikely to be scheduled during any particular segmentof the examination schedule. However, the availabilityand desires of the subjects do influence the compo-sition of the replicate sample. For instance, an exam-inee whose participation in an original examinationwas achieved only after repeated contacts by surveypersonnel is more likely to have been excluded froma replicate examination since it is unlikely that hewould have received an original examination during thefirst week and a half. The schedule of locations con-sidering time of year, sequencing of examinations,relation to other events which might make subjectsmore or less available, and other related aspects giveno obvious discriminatory factor. After examining theseand other relatively minor considerations there appearsto be no reason to believe that the subjects scheduledand examined during the first part of a stand differfrom those scheduled and examined during the latterportion of a stand with respect to the data gathered.

Another restriction on complete randomness in theselection of examinees for replicate examinations wasthe exclusion of those examinees who were “geograph-ically inconvenient” to the examination center. “Geo-graphically inconvenient” was arbitrarily defined as adistance of 30 miles or greater; although if conditionsdictated, exceptions were sometimes allowed. A pri-mary consideration in choosing a site for the exam-ination center was the centrality of the location inrelation to the sample segments (a segment is acluster of households). Since segments were drawnwith probability proportional to population, most seg-

63

Page 69: Skinfold Thickness of Youths12=17Years

ments were in relatively populated areas; and so theexamination center was also in or adjacent to arelatively populated area. Therefore, the subjects de-leted by this 30-mile restriction usually resided inrelatively less populated areas; so this reS&iCtiOn maycreate a bias in the replicate data if, in fact, charac-teristics and errors of concern differed by populationdensity. Even if differences did exist, the total effectof this restraint was not great since it excluded onlyapproximately 10 percent of the eligible examinees.There were other minor restrictions of medical andoperational nature imposed on the complete random-ness of the replicate sample, but they were not readilyassociated with large differences. Also they deleted atmost only 1-2 percent of the eligible examinees andfor these reasons are of small consequence.

Since the purpose of replicate examinations is togiveinformation about errors, the matter of concernbetween those excluded and those eligible for selectionis not the possible differences in the values of mess-urements but the possible differences in the errorsassociated with the measurements as shown by thediscrepancy between two measurements on the samesubject. It should also. be noted that although subjectsdid influence measurement errors [for some types ofexaminations], the environment, procedures, and exam-iners were also highly influential. The considerationof these additional influences causes a completelyrandom selection of subjects to be of somewhat lessconcern.

(Note: This concludes the material extracted fromSchaible’s paper.)

Evaluation of Residual MeasurementError in Skinfold Measurements

The residual error of measurement was estimatedfrom a set of 301 replicate examinations conducted, asoutlined below, during Cycle III of the HES.

Body measurements were taken on 6,768 youthsand these children comprised the HES Cycle 111sample.At 30 of the 40 locations (or stands) visited throughoutthe United States, replicate body measurements wereobtained on 301 children. That is, an average of 10youths were reexamined at each stand. Of the 301youths, 224 were reexamined by a technician other thanthe one initially measuring the youth, while the re-maining 77 were reexamined by the same technician.All together during the 4 years, 11 technicians partic-ipated in replicate measurements for this phase of thequality control program.

Tabie I presents the percentage of total exami-nations performed by each technician and the percent-ages of intra-examiner and inter-examiner replicatesin which the 11 technicians were involved.

The table below indicates some possible sourcesof bias which may affect the analysis of replicate data.For example, assume technician number 9 was able toreplicate his own measurements very well, but hisreadings were quite different from those of the otherexaminers. Obviously, his results would be overrepre-sented in the replicate analysis because he examinedonly 11.3 percent of all youths in the actual survey, butdid 16 percent of the intra-examiner replicate examina-tions and 13.3 percent of the inter-examiner replicateexaminations. Because of this technician’s overrepre -sentation in the replicate study, the distribution ofintra-examiner differences would cluster closer tozero than it really should have since this examinerself-replicates well. On the other hand, the inter-examiner distribution of differences would be consider-ably more skewed than it should have been since thistechnician does not agree well with the other technicians’measurements. Similar discrepancies are obvious forother technicians. An example of an opposite effect tothat cited above is technician number 2, who did only2.7 percent of the intra-examiner replicate measure-ments and 10.2 percent of the inter-examiner replicate

Table 1. Percentage of regular and replicate examinations performed by each techni.ci.an

Replicate examinations

Technician number Percentage of regularCycle III examinations

Percentage of intra - Percentage of f.nter -

examinations examinations

1 1.32 1!:: 1%;

22.8 2::; 21.4: 6.1 4.0 2.7c la G in-7 16.7

Ad. > -“. . ----

6.1 5.3N

1%: 2::: 16.411.3 16.0 1:.;

2:: ::; 3:6

64

Page 70: Skinfold Thickness of Youths12=17Years

measurements, but did 13,4 percent of all examinationsin Cycle III.

Thus, the various combinations of observers forthe inter-examiner replicates and the proportions ofintra-examiner replicates were not controlled so as tobe balanced among the observers. In the survey properthe examinations were similarly not proportionatelydistributed among the observers—an imbalance causedby the variation in the length of time the various tech-nicians were associated with the survey,

The foregoing indicates that the distribution ofnumbers of replicate examinations done by each tech-nic ian is not the same as the distribution of the totalnumber of survey examination done by each in Cycle111. This is one of the inherent problems of the presentreplicate data, and limits to some extent implicationsto the survey as a whole. Nevertheless, the readershould be aware of the many problems confronting thosewho conduct large-scale health surveys23 and in thiscontext, the present systematic approach to the col-lection of replicate body measurement data is adequate.

Results of

Replicate Examinations

The absolute differences between the first and thesecond examinations were computed for each child oneach of the five skinfolds of interest and the resultsare presented below.

Inter-Examiner Differences

There were 224 youths reexamined by a technicianother than the one who did the initial examination. Thedistributions of absolute differences between the findingsof the two examinations are shown in table II.

For four of the five skinfolds the modal differencewas 0.5 mm.; for the medial calf the mode was 1.0 mm.The largest mean differences were for the suprailiac(2.43 mm.) and medial calf (2.41 mm.); the triceps, at1.89 mm., was lower than these, but still higher thanthe midaxillary (1.33) or subscapular (1.34) skinfolds.The median differences for these latter two were lessthan for the other three.

A widely used measure of replicability is the sta-tistic cm, the technical error of measurement defined

nas .?.= ~d2 . This assumes that the distribution

nof replicate differences is normal and that the errorsof all pairs can be p@ed. The results of the calcula-

tions of this statistic are shown in table H. As ex-

pected, the largest values belong to the suprailiac andmedial calf skinfolds, the triceps displaying an inter-mediate value and the subscapular and midaxillary sitesthe smallest.

Triceps: 1.89Subscapular: 1.53Midaxillary: 1.47

Suprailiac: 2.45Medial calfi 2.44

This comparison is somewhat misleading since thesuprailiac and medial calf are the largest of the fiveskinfold measurements and have the greatest variance(see tables 1-5). On the other hand, the midaxillaryand subscapular akinfolds are highly correlated andhave similar distributions since both are trunk meas-urements. By expressing the technical error relativeto the appropriate mean, a coefficient of variation

(i.e., a measure of s’efotive error) is obtained. Thus,technical error

coefficient of variation = x 100average measurement

Since,. in Cycle III, the average values for theseskinfolds over all ages and sexes were:

Triceps: 12.25Subscapular: 9.97Midaxillary: 8.47

Suprailiac: 13.93Medial calfi 14.38

the following are the coefficients of variation:

Triceps: 15.44Subscapular: 15.37Midaxillary: 17.39

Suprailiac: 17.67Medial calti 17.09

Viewed in terms of relative error the suprailiacand medial calf skinfolds continue to show the poorestreplication. However, the midaxillary also falls intothis category.

lntra-Examiner Differences

A similar analysis was also conducted for the 77

youths reexamined by the same technician. The dis-tributions of differences are shown in table 111.

Here, the triceps measurement displays the small-est mean difference and, aa before, the medial calfand suprailiac the largest. These two also have thelargest median differences, the suprailiac being thehighest of the five. The technical error of the tricepsis the least of the five:

Triceps: 0.80Subscapular: 1.83Midaxillary: 2.08

Suprsiliac: 1.87Medial calf: 1.44NOTE: The list of references follows the text.

65

Page 71: Skinfold Thickness of Youths12=17Years

Table II. Distribution of inter-examinerdifferences between the initial and the rc. ..A. -––. .—-A. ––— e.. – -L. x.–—. .,.. . . . .. .

.acace exammauons zor me rave sm.nrows

iiiiir=e~‘bsO1ute Difference Triceps skinfold

E

3ubscapularskinfold Midaxillary skinfoli Suprailiac skinfold Medial calf skinfold

?requency1 Percent Frequencyl I Percent Frequencyl ?mquencyl I Percent?ercent

:: 2::;0.0

01::!

X 9.41622 :::18

:::: 4.94 1.812 ::;2 0.92 0.9

0.9: 0.00 0.00 0.00 0.0

: i:o 0.01 0.5

: t!1 0.50 0.00 0.00 0.0

10.4:; lj.:

o1!::

:(34) 14.410.4

16 7,23.6

1:12 ;:;3 .

: ::!44 1:81

::;: 0.90 0.00 0.0

: :::0.9

: 0.00.5

: 0.01 0.50 :.g11 0:5

0.0

:?:-0.9

;::

2:53.0

t:4.5

i:6.06.5

;::8.0

;:5

1:::10.5;:,g

12:012.513.013.516.0

48611

4:1715114

:0532

:0111

:10

:0000

21.427.2:.:

1;::

u1.81.32.70.0

:::0.90.4

$;0.40.40.4

:::0.40.0

$80.00.00.00.0

40640

:;

1796331612001

:0

:10000000

17.928.60.0

1$::1+.;

4:02.71.3

;::

::i0.90.00.00.40.40.00.0

:::0.40.0

%:0.00.00.00.0

:?o

3:24(26)1912161:

4

:44;

211

:10011110

11.316.70.0

1!:!10.88.65.4

;::

:::

;:;1.81.8

i :0.90.50.5

::;0.50.0

:::0.50.50.50.0

I

224 100.0 \ 224 ] 100.0Sample size, nMean difference,

z, in mm.Median in mm.Mode in um.Zd2Zd2f2n

224 100.0 222 100.0

1.891.5

160;:23.58

*.;4 1.331.0

105!:? 97!;;2.35 2.17

2.431.50.5

2664.86.00

2.411.5

264$ !5.97

lNumberof replicate examinations exhibiting indicated differences.2Such differencesmay be caused by failure of technicians to round measurement to nearest half-millimeteror by a miscoding

error undetected during imputation.

Table III. Distribution of intra-examinerdifferencesbetween the initial and the reDlicate examinations for the five skinfolds

Triceps skinfold ubscapularskinfold sdial calf skinfold:idsxillaryskinfold :uprailiacskinfold$yf;tc

ence, d,ill IMU

1ifferencesqu;~ed,

kequencyl Percent Percent FrequencylFrequenty1 Frequencyz Frequenty1Percent Percent Percent

0.00.51.01.52.0

:::

M4.55,05.56.07.0

18::19.023.0

27.335.116.99.15.2

N0.01.3

:::0.00.0

8::0.00.00.0

100.0

37.728.614.3;.:

:::1.31.3

:::0.0

iii0.00.0

$:

100.0

31.231.218.25.26.51.3

M:.:

2:60.00.0

:::0.00.01.3

10.416.916.914.39.1

?::2.66.50.05.21.31.3

i:

:::0.0

Sample size,nMean difference,a,in w

Median in um.Mode in mm

;%/2n

77 77 77 I 100.0 77 100.0 77 100.0

0.780.5

9:::0.64

1.050.5

51::!3.34

1,100.5

66::;4.31

1.971.5

52;:~3.44

1.441.0

330::2.15

1Number of replicate examinationsexhibiting indicateddifferences.

Page 72: Skinfold Thickness of Youths12=17Years

Computation of coefficients of variation accentuatesthe intra-examiner precision in triceps measurement.The coefficients of variation are as follows:

Triceps: 6.51Subscapular: 18.33Midaxillary: 24.51

Suprailiac: 13.45Medial calfi 10.36

The triceps measurement obviously has the lowestcoefficient of variation of the five skinfold measure-

ments under consideration, while the midaxillary skin-fold exhibits the worst replicability.

Within each skinfold, the significance of differ-ences between the intra- and inter-observer errors ofmeasurement were tested by computing the F ratiosof their squares. The results were as follows:

Triceps: 4.20Subscapular: 1.42Midaxillary: 1.29

Suprailiac: “1.73Medial calfi ‘2.72

Tests of these at the .01 level (to keep the overallerror rate below 5 percent) showed that the triceps,suprailidc, and medial calf were those of the five inwhich the inter- and intra-examiner technical errorsdiffer significantly. That is, agreement was signifi-cantly better when the same observer replicated theinitial measurement. For the other two skinfolda, theerror associated with two observers was no greaterthan the intra-observer error.

These findings indicate that error in skinfold meas-urement is related to both the site and number ofobservers utilized. The measurement of the thicknessof the triceps skinfold involves highly individualtechniques, probably related to the precise spot overthe muscle, the manner in which the fold is “pickedUp,11 and the pointat which the caliper faces make

contact with the skin. In addition, although a formallyanalyzed study was not conducted, a clinical impressionwas formed in the training sessions that the precisesite chosen for measurement was more critical in thetriceps region than in the others (presumably the sub-cutaneous fat varies in thickness more in the triceps

region as one strays from the exact site—i.e., aroundthe circumference of the arm —tbsn in the other regions).A single observer will become quite consistent in termsof his or her own technique, and self-replication willbe quite high.

Suprailiac and medial calf replication is also re-lated to observer number. The suprailiac skinfold dis-plays, in some cases, extremely high values in certainindividuals, and slight differences in individual tech-niques may cause aignificsntly poorer replicates be-tween observers.

The medial calf skinfold is technically the mostdifficult of the five to measure. In addition, as with thesuprsiliac, its values may be quitelarge.Thereforeit is subject to significant inter-observer error.

On the other hand, such individualized techniquesare not as important for the subscapular and midaxil-lary skinfolds since the adipose layer in these regionsis thinner and more uniform in thickness than in thearm, The associated error is more likely to be ran-domized and not to be so strongly affected by “examiner-specific” factors.

Conclusions

From the above, some conclusions may be drawnrelative to the error of measurement associated withthe HES. The median error for all skinfolds is 1.0 to1.s mm. This error, though absolutely small, is rela-tively quite high in view of the usual thickness of skin-folds encountered. In addition, quite large errors canoccur, replicate differences of 12.5 mm. being observedfor the triceps, 10.5 mm. for the subscapular and mid-sxillary folds, 13.5 for the suprailiac, and 16.0 for themedial calf fold. These errors remain as residualsdespite the careful quality control exercised through-out Cycle III. The meaning of such errors may beevaluated only in light of the fact that the measurementof body fat is of considerable biomedical import and,in many cases, skinfolds provide tbe only estimatesavailable.

With well-trained and supervised observers, theresidual errors of measurement are the same for thesubscapular and midsxillary skinfolds regardless ofwhether one or several observers are utilized. Such isnot the case for the triceps, suprailiac, or medial calf,however, because the residual is significantly less whenonly one observer does the measuring. On the other hand,the possibility of systematic errors is greater withonly one observer, leading to a potentially systematicbias in the distributions.

In a longitudinal study, a single observer is alwayspreferable. The major purpose of such a design is todetermine change in individuals over time. A singleobserver will provide more consistent readings andtherefore a more accurate estimate of change. How-ever, since use of a single observer increases thepossibility of systematic bias, the reliability of longi-tudinal studies is reduced for estimates of the distri-bution of absolute values in the general population.

In a cross-sectional study, multiple examiners arepreferable so far as the subscapular and midaxillaryskinfolds are concerned. Not only are residual errorsof measurement the same regardless of whether oneor several observers are used, but also the systematicbias introduced by use of a single observer will beeliminated.

For cross-sectional studies involving the otherthree skinfolds, the situation is more complex. If

67

Page 73: Skinfold Thickness of Youths12=17Years

the purpose is to estimate their distributions in apopulation, multiple examiners will provide better es-timates since systematic error will be more likely tobe reduced.

If the purpose is to make comparisons betweengroups, then the design of the study, based on consider-ations of all factors, must reconcile two opposingproblems:

(1) Multiple examiners will increase the variability ofthe distribution because of the inclusion of inter-examiner errors of measurement.

(2) Single-examiner measurements will result in avariance more comparable to the true value for thepopulation. However, since a single observer maymeasure different kinds of individuals in a system-atically different way, new problems of bias maybeintroduced.

—ooo —

68

Page 74: Skinfold Thickness of Youths12=17Years

Series 1.

Series 2,

Series 3.

Series 4.

Series 10.

Series 11.

Series 12.

Series 13.

Series 14.

Series 20.

Series 21.

Sevies 22.

OUTLINE OF REPORT SERIES FOR VITAL AND HEALTH STATISTICS

Originally Public Health Service Publicatwn No. 1000

Programs and collection pvocedures.— Reports which describe the general programs of the National

Center for Health Statistics and its offices and divisions, data collection methods used, definitions,and other material necessary for understanding the data.

Data evaluation and methods research. —Studies of new statistical methodology including: experi-mental tests of new survey methods, studies of vital statistics collection methods, new analyticaltechniques, objective evaluations of reliability of collected data, contributions to statistical theory.

Analytical studies. — Reports presenting analytical or interpretive studies based on vital and healthstatistics, carrying the analysis further than the expository types of reports in the other series.

Documents and committee reports. — Final reports of major committees concerned with vital andhealth statistics, and documents such as recommended model vital registration laws and revised birthand death certificates.

Data from (he Health Interview SuYvey. —Statistics on illness, accidental injuries, disability, use ofhospital, medical, dental, and other services, and other health-related topics, based on data collectedin a continuing national household interview survey.

Data from the Health Examination Survey. — [)ata from direct examination, testing, and measure-ment of national samples of the population provide the basis for two types of reports: (1) estimatesof the medically defined prevalence of specific diseases in the United States and the distributions of

the population with reqpect to physical, physiological, and psychological characteristics; and (2):malysis of relationships among the various measurements without reference to an explicit finiteuniverse of persons.

Data from the Institutional Population Surveys. — Statistics relating to the health characteristics ofpersons in institutions, and on medical, nursing, and personal care received, based on nationalsamples of establishments providing these services and samples of the residents or patients.

Data from the Hospital Discharge Survey.— Statistics relating to discharged patients in short-stayhospitals, based on a sample of patient records in a national sample of hospitals.

Data on health resources: manpower and facilities. — Statistics on the numbers, geographic distri-bution, and characteristics of health resources including physicians, dentists, nurses, other healthmanpower occupations, hospitals, nursimz homes, and outpatient and other inpatient facilities.

Data on mortality. —Various statistics on mortality other than as included in annual or monthlyreports —special analyses by cause of death, age, and other demographic variables, also geographicand time series analyses.

Data on natality, marriage, and divorce. — Various statistics on natality, marriage, and divorce otherthan as included in annual or monthly reports— special analyses by demographic variables, alsogeographic and time series analyses, studies of fertility.

Data from the National Natality and Mortality Surveys. —Statistics on characteristics of births anddeaths not available from the vital records, based on sample surveys stemming from these records,inducting such topics as mortality by socioeconomic CIP~s, medical experience in the last year oflife, characteristics of pregnancy. ~’tc.

For a list of titles of reports published in these series, write to: Office ui’ Informationv

National Center for Health Statistics

/ublic Health Service, HRA

Rockviile, kid, 20852

/

Page 75: Skinfold Thickness of Youths12=17Years