Total Serum Cholesterol ValuesTotal Serum Cholesterol Values of Youths 12-17 Years United States...
42
Data from the NATIONAL HEALTH SURVEY Total Serum Cholesterol Values of Youths 12-17 Years United States Serum cholesterol values are presented and discussed by age, sex, race, and socioeconomic level of youths 12-17 years of age in the United States, 1966-70. DHEW Publication No. (H RA) 76-1638 U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service Health Resources Administration National Center for Health Statistics Ro,ckville, Md. May 1976 Series 11 Number 156
Transcript of Total Serum Cholesterol ValuesTotal Serum Cholesterol Values of Youths 12-17 Years United States...
Data from theNATIONAL HEALTH SURVEY
Total Serum
Cholesterol Values
of Youths 12-17 Years
United States
Serum cholesterol values are presented and discussed by age, sex,race, and socioeconomic level of youths 12-17 years of age in theUnited States, 1966-70.
DHEW Publication No. (H RA) 76-1638
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFAREPublic Health Service
Health Resources AdministrationNational Center for Health StatisticsRo,ckville, Md. May 1976
Series 11Number 156
Library of Congress Cataloging in Publication Data
Main entry under title:
Total serum cholesterol values of youths 12-17 years, United States.
(Vital and health statistics: Series 11, Data from the National Health Survey, Datafrom the health examination survey; no. 156) (DHEW publication no. (HRA) 76-1 638)
Includes bibliographical references.Supt. of Dots. No.: HE 20.6209:11/1561. Youth–Health and hygiene–United States–Statistics. 2. Blood cholesterol.
3. Health surveys–United States. I. Levy, Paul S. II. Series: United States. National Centerfor Health Statistics. Vital and health statistics: Series 11, Data from the National HealthSurvey, Data from the health examination survey; no. 156. III. Series: United States, Dept.of Health, Education, and Welfare. DHEW publication; no. (HRA) 76-1 638) [DNLM:1. Cholesterol–Blood. 2. Blood chemical analysis-In adolescence. W2 N148vk no. 156]RA407.3.A347 no. 156 [RA564.5] 312’.0973s [612’.12] ISBN 0-8406-0051-875-619260
NATIONAL CENTER FOR HEALTH STATISTICS
DOROTHY P. RICE, Director
ROBERT A. ISRAEL, Deputy Direckx-JACOB J. FELDMAN, Ph.D., Associate DirectorforAna@is
GAIL F. FISHER, Associate Director for the Cooperative Health Statistics SystemELIJAH L. WHITE, Associate Director for Data Systems
GEORGE P. FAILLA, Associate Director for ManagementPETER L. HURLEY, Associate Director for Operations
JAMES M. ROBEY, Ph.D., Associate Director for Program DeveibpmentALICE HAYWOOD, Information Officer
DIVISION OF HEALTH EXAMINATION STATISTICS
ARTHUR J. McDOWELL, Director
JEAN-PIERRE HABICHT, M.D., SC.D., Special Assistant to Director
PETER V. V. HAMILL, M.D., Medical Adviser
JEAN ROBERTS, Chiej Medical Statistics Branch
ROBERT MURPHY, Chief Survey Planning and Development Branch
Vital and Health Statistics-Series 1l-No. 156
DHEW Publication No. (HRA) 76-1638
Library of Congress Catalog Card Number 75-619260
CONTENTS
Introduction . . . . . . . .
Method . . . . . . . . . . .Race . . . . . . . . . . .Blood Specimen and Total
Findings . . . . . . . . . .Age and Sex . . . . . . .
. . . . . . . . . . . . . .. . . . .
. . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .Serum Cholesterol Determination
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .Geographic Region, Age, and Sex . . . . . . . . . . . . .Race, Age, and Sex . . . . . . . . . . . . . . . . . . . .Annual Family Income, Education of Parent, Age, and Sex
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . .Age and Sex . . . . . . . . . . . . . . . . . . . . . . . .Race . . . . . . . . . . . . . . . . . . . . . . . . . . . .Annual Family Income . . . . . . . . . . . . . . . . . .Other Studies . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . .
Listof Detailed Tables . . . . . . . . . . . . . . . . . . . .
Appendix I. Technical Notes on Methods . . . . . . . . . .The Survey Design . . . . . . . . . . . . . . . . . . . . .Some Notes on Response”Rates . . . . . . . . . . . . . .Parameter and Variance Estimates . . . . . . . . . . . . .Standards of Reliability and Precision . . . . . . . . . . .Imputation . . . . . . . . . . . . . . . . . . . . . . . .
Appendix II. Demographic Variables . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Appendix III. Techniques of Measurement and Quality ControlMeasurement . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
The HES Blood-Drawing Technique and Its Historical Development . . . . . .Cholesterol Chemistry Determinations . . . . . . . . . . . . . . . . . . . . .
Monitoring Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Methods of Taking Replicate Measurements . . . . . . , . . . . . . . . . . .Results of the Replicate Study for Cholesterol . . . . . . . . . . . . . . . . .Data Handling Verification . . . . . . . . . . . . . . . . . . . . . . . . . .
Page
1
223
33445
679
1010
11
13
262627282929
30
3131313232333335
SYMBOLS
I Data not available--------------------------------------- --- ICategory not applicable ------------------------------ . . .
I Quantity zero --------------------------------------------- - II Quantity more than Obut less than 0.05---- 0.0
Figure does not meet standards ofreliability or precision ---------------------------- *
TOTAL SERUM CHOLESTEROL VALUESOF YOUTHS 12-17 YEARS
Paul S. Levy, SC.D., Peter V. V. Hamill, M.D., M.P.H., Felix Heald, M.D., and Michael Rowlanda
INTRODUCTION
This report of serum cholesterol values asdetermined from blood samples of youths 12-17years of age in the United States is one of aseries of reports presenting findings from CycleIII of the Health Examination Survey (HES).The means and selected percentiles of totalserum cholesterol values are examined here bysex, age, race, family income, education, andgeographic location in the United States. Asdescribed in a detailed report of its general planand operation, 1 the Health Examination Surveyis conducted in a succession of cycles.
Cycle I of the HES, conducted from 1959 to1962, obtained information on the prevalence’ ofcertain chronic diseases and on the distributionof a number of anthropometric and sensorycharacteristics in the civilian, noninstitu-tionalized population of the continental UnitedStates aged 18-79 years. The detailed plan ofCycle I has been described,z and most of theresults are published in other reports in Series 11of Vital and Health Statistics.
Cycle II of the HES, conducted from July1963 to December 1965, involved selection andexamination of a probability sample of noninsti-tutionalized U.S. children aged 6-11 years. Thisprogram succeeded in examining 96 percent ofthe 7,417 children selected for-the” sample. The
‘AssociateProfessorof Biometry,Universityof Wlnok,School of Public Health; Medlcaf Adviser, Children and YouthPrograms, Division of Health Examination Statistics; Professor ofPediatrics, D]rector, Division of Adolescent Medicine, Universityof Maryland, School of Medicine; Analytical Statistician,Division of Health Examination Statistics, respectively.
examination had two focuses: (1) factors relatedto healthy growth and development as deter-mined by a physician, a nurse, a dentist, and apsychologist and (2) a variety of somatic andphysiologic measurements performed by spe-cially trained technicians. The detailed plan andoperation of Cycle II and the response resultsare described in Vital and Health Statistics,Series l-Number 5.3
HES Cycle III, conducted from March 1966to March 1970, was essentially an agewise exten-sion of Cycle II into adolescence. As describedin detail in “Plan and Operation of a HealthExamination Survey of U.S. Youths, 12-17Years of Age,”h Cycle III was more similar toCycle II than to Cycle I not only in form,content, and style but also in having its majoremphasis on factors of “normal” growth anddevelopment rather than on chronic diseases.These analyses on “normal” growth and devel-opment of adolescents have been well underwaysince 1970, and some of the results from thebattery of body measurements have already
‘ been published,s ‘T as have the initial results ofthe hematocrit findingsg and the serum uric acidvalues.g
The present report of cholesterol values is thethird in the series presenting findings from thesample of blood drawn from each youth. Noblood specimens were obtained from children inCycle II, but specimens were obtained from theadults in Cycle I.
The information from the analysis of blood isintended for use as much needed reference databoth for clinical and for epidemiologic estimates”of variation in a well-defined population. It also
enables examination of another aspect of growthand development during adolescence. The proce-dure by which the Division of Health Examina-tion Statistics obtained advice on the selectionof the analyses to be made on the blood speci-mens and the cooperative arrangements whichwere subsequently made with the Lipid Stand-ardizatipn Laboratory of the Center for DiseaseControl (CDC) have been described elsewhere.A
The authors and the Division of HealthExamination Statistics are grateful for the tech-nical advice given by Dr. Gerald Cooper, Dr.Myron Kuchmak, and Dr. Alan Mather and fortheir administrative assistance in arranging thelaboratory determinations that were performedby the Lipid Standardization Laboratory atCDC, and to Mrs. Margie Sailors for coordina-tion and verification of the laboratory data andits transmission.
METHOD
At each of 40 preselected locations (seeappendix I for sample design) throughout theUnited States, the youths were brought to thecentrally located mobile examination center foran examination that lasted about 3% hours. Sixyouths were examined in the morning and six inthe afternoon. Except during vacations, theywere transported to and from school and/orhome.
When the youths entered the examinationcenter, their oral temperatures were taken, and acursory screening for acute illness was made; ifillness was detected, the youth was sent homeand reexamined later. The examinees changedinto gymnasium-t ype shorts; cotton sweat socks;a terry-cloth robe; and, for the girls, a light,sleeveless topper. All six then proceeded todifferent stages of the examination, each onefollowing a different route. The 3?4-hour exami-nation was divided into six 35-minute timeperiods, each consisting of one or more detailedexaminations at a designated station. At the endof each period, the youths rotated to other
stations, so that at the end of 3% hours eachyouth had been given essentiality the sameexaminations by the same examiners, but in a
different sequence. Four of these examinationtime periods were allocated to examinations bya pediatrician, a dentist, and a psychologist,band the other two were allocated to a group ofexaminations performed by highly trained tech-nicians. This last group of examinations con-sisted of X-rays of the chest and hand-wrist,hearing and vision tests, measures of respiratoryfunction, a 12-lead electrocardiogram, a sub-maximal exercise tolerance test on a treadmillwith chest leads to a continuous electrocar-diogram, a battery of body measurements, gripstrength measurements, examination of bloodand (on girls only) urine cultures for bacteria,and a privately administered health behavior andhealth attitude questionnaire.
Race
Race was recorded as “white,” “Negro,” and“other races’” (see appendix II). In Cycle III,white youths constituted 84.74 percent of thetotal youths examined; Negro youths, 14.76percent; and youths of other races, only 0.50percent. In Cycle II, white children constituted85.69 percent of the examined subjects andNegro children, 13.86 percent. (The differentialresponse rate by age, sex, and race is analyzedand discussed in appendix I. The increasedproportion of Negro subjects in Cycle 111was
bThe entire examination by the psychologists consisted of
two consecutive time periods (70 minutes). Two psychologistsperformed identical examinations simultaneously at separatestations.
“rhe same classification scheme as used in the 1960 censuswas employed here. As described in the previously mentionedreport on the operation of HES Cycle 111,4 this information wasobtained at the initial household interview by the U.S. Bureau ofthe Census fieldworker. The accuracy of the information waschecked at the subsequent home visit by the highly experiencedrepresentative from HES and again at the examination in the@roller. A final record check by birth certificate turned up onlyseven inconsistencies, and these were mostly pertaining to thecategory “other races,” Hence, the possible extent of misclassifi-
cation of the race variable, as described, is so minimal that itcould have no effect on the data analyzed in this report.However, when comparing the present HES findings to those ofother variouply defined racial groupings in the world, the degreesof genetic admixture, as first discussed by Herskowitsl Oin 1928and later by Glass and U, 11 by Robertsjl 2 Roberts andHioms, la and by Reed14 should be taken into consideration.
due to their better response rate–the overallNegro response rate wa~ 96.6 percent and theoverall white response rate was 89.1 percent.) Asin Cycle II, because so few youths of “otherraces” were part of the sample, data for themhave not been analyzed as a separate category.Whenever data are analyzed independently of aclassification by race, however, data for theseyouths are included.
Blood Specimen and Total SerumCholesterol Determination
The analyses for total serum cholesterol (i.e.,both free cholesterol and esterified cholesterol)were performed for the Health ExaminationSurvey at the Lipid Standardization Laboratoryof the Center for Disease Control; Public HealthService, Atlanta, Ga.. There, a semiautomatedprocedure was employed using the Abell-Kendallmethod,l 5 details of which are presented inappendix III. This method, with a specificity y fortotal cholesterol of 99 percent, is considered anaccurate measure of total cholesterol. 1s>1G Ithas the additional advantage of being used inmost of the controlled studies on the role ofcholesterol in heart disease, as it yields highcomparability y.
The development of the technique used bythe HES for obtaining and processing the bloodspecimen is also described in appendix III. Asdescribed in the two previous reports,s ‘g themicrohematocrit was the only complete labora-tory determination performed on the bloodspecimen directly in the examination trailers.The bulk of each blood specimen, after prelimi-nary laboratory preparation, was properlyseparated into its various subsamples and packedinto specially devised styrofoam containers forshipment via air freight to either the Immuno-genetics Laboratory of The Johns Hopkins Uni-
versity, Baltimore, Md., or the Center for DiseaseControl in Atlanta, Ga.
A frozen specimen for the total serum choles-terol was shipped directly to the Clinical Chem-istry Section of the Center for Disease Controlfor analysis. A specimen of clotted blood wassent directly to the Venereal Disease ResearchLaboratory of the CDC for analysis.
FINDINGS
Age and Sex
The estimated number and percent distribu-tion of male and female youths 12-17 years ofage falling into each of 19 groups according totheir levels of cholesterol are shown in tables 1and 2, respectively. The patterns shown in thesedistributions are discussed below in terms of afew summary statistics.
The estimated mean cholesterol levels foreach of the 12 age-sex classes are shown in table3 and figure 1. Among male youths 12-17 yearsof age, mean cholesterol levels were highest in12-year-olds (181.3 mg/100 ml), decreased inthe next two age groups (174.7 mg/ 100 ml in13-year-olds and 169.9 mg/100 ml in 14-year-olds), and remained relatively stable in the nextthree age groups. The mean cholesterol levels offemales aged 12-16 ranged from 175 to 178mg/ 100 ml; but for 17-year-olds, however, themean level was 184.5 mg/ 100 ml—indicating arise of more than 6 mg/100 ml from 16 to 17years of age. With the exception of 12-year-olds,
200
F=
.- - —------------ —------------
— Male---- Female
“12 13 14 15 16 17
AGE IN YEARS
Figure 1. Mean cholesterol levels for youths 12-17 years of age,
by age and sex: United States, 1966-70.
3
females in each age group had higher meancholesterol levels than males in the same agegroup.
Percentiles of the cholesterol distributions areshown by age for male and female youths (table3 and figure 2). For each age-sex group, themean cholesterol level was always greater thanthe median, which indicates that the cholesteroldistribution was skewed to the right (positiveskewness). This skewness was most pronouncedin 17-year-old females for whom the meancholesterol level was 5.3 mg/ 100 ml greater thanthe median. The age patterns described abovefor the mean were also apparent when percen-tiles at both the upper and lower ends of thedistribution were examined.
Geographic Region, Age, and Sex
Mean cholesterol levels are shown by age andsex for the four geographic regions defined inthe Health Examination Survey in table 4 andfigure 3. In general, mean cholesterol levels werehighest in the Northeast, and there were few, if
any, differences in mean cholesterol levelsamong the other three regions.
Race, Age, and Sex
The estimated number and percent distribu-tion of white and Negro youths falling into eachof 19 groups according to their cholesterol levelsare shown, respectively, in tables 5, 6, 7, and 8.Patterns observed in these distributions are dis-cussed below in terms of a few summary statis-tics.
The main cholesterol patterns observed abovefor the entire sample of U.S. youths aged 12-17. . .——.years were similar to those observed for whiteand Negro youths separately (table 9 and figures4 and 5). Mean cholesterol levels in white maleswere highest for 12-year-olds (180.2 mg/100 ml)and decreased with age to a low of 166.7mg/100 ml for 16-year-olds. For white females,there was little change with age in mean choles-terol levels except for a sharp increase from177.7 mg/100 ml for 16-year-olds to 184.2mg/ 100 ml for 17-year-olds. Mean cholesterol
280
240
120
Male
0112 13 14 15 16 17
AGE IN YEARS
280
tFemale
u! I
012 13 14 15 16 17
AGE IN YEARS
Figure 2. Selected estim6ted percentiles of the cholesterol distribution in youths aged 12-17 years, by age and sex: United .St8teS,
1966-70.
4
190t Male
_ Northeast
J---Midwest
___ Seuth
---- West
o~12 13 14 15 16 17
AGE IN YEARS
190
t
Female
160--... ~,--:>...<,’%... ..% “.. /’ .7-/
------ ,x. /“‘\\ ~. #.“,’ . . ... ..~
170
‘6:d’12. 13 14 15 16 17
AGE IN YEARS
Figure 3. Mean cholesterol Ievelsof maleand femeleyouths 12-17years ofage, byageand geographic region: United States, l9~7O.
&Males
200
Females
200 I------------------------ ---------------- -- ---e --------
wd
--------------- -------
d
~z&-+Ags
5!8~ ~ 100- :: 100 -
Zz00$Z
$z
White _ White
Negro2
---- —--- Negro
0 I , , 0 t t 1 ,12 13 14 15 16 17 12 13 14 15 16 17
AGE IN YEARS AGE IN YEARS
Figure 4. Mean cholesterol levels of male and female youths 12-17 years of age, by age and race: United States, 1966-70.
levels in Negro males were highest for 12-year-olds (187.2 mg/100 ml) and decreased with ageto a low of 173.3 mg/100 ml for 16-year-olds.
At every age except 12 years, both white andNegro female youths had higher mean choles-terol levels than their male counterparts. At age12, however, the mean levels were higher forboys of both races than for girls.
In addition, Negro male and female youthshad higher mean serum cholesterol levels thantheir white counterparts at every age except 12,with interracial differences being greater formales than for females. The average differences
between the mean cholesterol levels of whiteand Negro youths over the six age groups was7.8 mg/100 ml for males and 2.7 mg/100 ml forfemales. In general, interracial differences weregreater at the upper end of the cholesteroldistribution than at the lower end (figure 5).
Annual Family Income, Educationof Parent, Age, and Sex
Mean cholesterol levels are given by annualfamily income, age, and sex in table 10 andeducation of parent, age, and sex in table 11. No
240t
Males
I-------------
..z~
3~ 200
g------
3 ----------------- ----- SO*
~
~ 160
z~
----------------------- ----UJ20a 120
I _ White----Negro
0’ 1 , 1 ,
12 13 14 15 16 17
AGE IN YEARS
[– y
Females24o
------ goth-.0-... ---------
~--- -. -----------
=z 200
~ ..--% ---z .-- ----------- ------ 5oth
~ P ‘----
~ 160Kw~
t-
/ Ioth- ---------- --- - -------
*w.
----------
I — White
---- Negro
01 t 1 , 1 1
12 13 14 15 16 17
AGE iN YEARS
Figure 5. Selected estimated percentiles of the cholesterol distribution in male and female youths 12-17 years of age, by age and race:
United States, 1966-70.
consistent relationship was observed betweencholesterol levels and either annual familyincome or education of parent within individualage groups. However, when all age groups werecombined, the mean cholesterol levels averaged3.4 mg/ 100 ml higher for male youths and 6.8mg/ 100 ml higher for female youths whosefamilies had annual incomes of $10,000 or morethan for those youths whose families had in-comes of less than $3,000 per year (table 12).Similar patterns were observed in the relation-ship between mean cholesterol levels and paren-tal education.
Mean cholesterol levels are shown separatelyfor white and Negro youths by annual familyincome, age, and sex in table 13 and by educa-tion of parent, age, and sex in table 14. When allage groups were combined (table 12), a generalincrease in mean cholesterol with increase infamily income was observed for both white andNegro youths. Likewise, both white and Negroyouths whose parents had 12 years or more ofeducation generally showed higher mean choles-terol levels than those whose parents had lessthan 12 years of education. When contrastingthe socioeconomic extremes, youths of aIl agesin the highest socioeconomic groups had highercholesterol levels than those in the lowest socio-economic groups.
DISCUSSION
The role of cholesterol as a risk factor inincreased morbidity and mortaIity from cardio-vascular disease has been an issue for decades.Evidence of its centrzd role in the developmentof atherosclerosis comes from three sources.First, certain disease states in man involvingabnormally high serum cholesterol levels are alsoassociated with accelerated atherosclerosis. Ex-amples of such disease states are hyperlipopro-teinemias (Fredrickson Types II, III, and IV) andpoorly controlled diabetes mellitus, both entitiesknown to be associated with accelerated athero-sclerosis.17 Second, “epidemics” of atheroscle-rosis currently occurring in the United Statesand the Scandinavian countries are unknown inpopulations where serum cholesterol levels arelow. Although high levels of serum cholesterol inepidemiological studies are not always associatedwith high rates of atherosclerosis, low levels ofserum cholesterol always appear to be associatedwith low levels of atherosclerosis. 18 Third, it h=been known for many years from animal experi-ments and human autopsy data that esterifiedcholesterol is the principal lipid constituent ofthe fatty streak and the fibrous plaque.19 ~zO
Despite these facts, there has been consider-able debate over the role cholesterol plays in the
6
pathogenesis of atherosclerosis. The debate hasranged from advocating cholesterol as the majorand primary determinant in the development ofatherosclerosis to the other extreme of describ-ing its role as only a secondary or facilitatingone. Current opinion seems to implicate elevatedlevels of serum lipids as one of a number ofmajor risk factors such as hypertension, smok-ing, and diabetes mellitus.z 1 Nongenetic eleva-tion of serum cholesterol occurring in one of thetechnologically developed countries is usuallyassociated with a high-calorie diet characterizedby significant amounts of cholesterol and animalfat. To complicate the issue even more, recentevidence implicates hypertriglyceridemia as anindependent risk factor in the development ofischemic heart disease.zz Therefore, it is possi-ble that serum cholesterol may not be the onlylipid in the serum to signify higher risk for heartdisease in middle-aged adults.
It is important to note, however, that serumcholesterol may continue to be a useful indexdetermination for high &k for cardiovasculardisease for three reasons. First, the exact role ofhypertriglyceridemia as a predictor of risk incardiovascular disease is not as solidly based asthat of hypercholesterolemia. Second, choles-terol levels are not as affected by the immediatefed or fasted state as are the serum triglycerides.Third, serum triglycerides may reflect the cur-rent nutritional status in days or weeks, whereasserum cholesterol is more likely to be repre-sentative of lipid metabolism over a period ofmonths. Therefore the’ accurate measurementand interpretation of serum triglyceride levelsrequires immediate dietary knowledge which isnot as important when using serum cholesterolfor survey or screening purposes.
Much of our epidemiologic knowledge aboutthe clinical implications of elevated serum cho-lesterol and triglycerides in adult morbidity andmortality is relatively recent, and it is basedessentially on adult data such as the classicallongitudinal study conducted in Framingham,Massachusetts.z 1 There are no analogous longi-tudinal data gathered directly from adolescentsthat link elevated cholesterol in adolescence(whether it be of genetic and dietary origin orsecondary to other diseases such as hypothyroid-ism, diabetes mellitus, and the nephrotic syn-
drome) with future morbidity and mortality.There ‘ is, however, evident; from autop;ystudies that fatty streaking of the abdominalaorta, a precurser to atherosclerosis, can be wellestablished during adolescence and may acceler-ate more rapidly at adolescence than at anyother time,z 3 and that accelerated coronarydisease can be found (though uncommonly) inmen in their twenties and with increasing fre-quency in men in their thirties.z4 Thus, ifelevated serum cholesterol plays a significantcausal role in adult atherosclerotic disease, and ifit could be demonstrated that purposefullylowering these levels by dietary and/or othermeans significantly lowers the risk, then theadolescent period would not be too early toattempt its detection and control.
Age and Sex
The relationship between serum cholesterollevels and age is different in adolescent malesand females in the United States. Mean choles-terol levels in males declined from a high of181.3 mg/100 ml in 12-year-olds to 167.7mg/ 100 ml in 16-year-olds, with the biggestdifferences between adjacent age groups oc-curring between ages 12 and 13 (6.6 mg/100 ml)and between ages ,13 and 14 (4.8 mg/100 ml).From age 14 on, there appeared to be little ifany change in mean cholesterol levels amongmales. On the other hand, there was little if anydifference in mean cholesterol levels of femalesaged 12-16 years, with mean levels ranging from175 to 178 mg/100 ml. At age 17, however,there was a sudden increase of 6.4 mg/ 100 ml inmean levels over that observed from 16-year-oldfemales–178.l mg/100 ml vs. 184.5, respec-tively. Except at age 12, female youths in eachage group had higher mean cholesterol levelsthan their male counterparts (table 3).
Among adolescent males and females of allages, the distribution of cholesterol levels wasnot symmetric but skewed to the right (positiveskewness). In male youths 12-17 years of ageand in female youths 12-16 years of age, themean of the cholesterol distribution was 1.5 to3.0 mg/100 ml higher than17-year-old females, however,
the median. Inthe distribution
7
was much more skewed with the mean being 5:3mg/ 100 ml greater than the median.
These estimates for U.S. youths with respectto age and sex can be compared with findings ofother studies of cholesterol in adolescents. Astudy conducted in Tecumseh, Mich., k 1959-60involved an extensive medical examination of 88percent of the town’s entire population (approx-imately 9,800 persons at that time). Included inthis examination were serum cholesterol deter-rninations.z 5 The Tecumseh study representsmore than just a collection of persons since itsresults can be extrapolated to a specified popula-tion (i.e., that of Tecumseh, Mich., 1959-60).Unfortunately, because Tecumseh findings oncholesterol were presented for the combined agegroups 10-14 years and 15-19 years, a moredetailed comparison with HES results is notpossible. However, in both of the Tecumseh agegroups, females had higher mean cholesterollevels than males, a finding consistent with thatof the HES.Zs
In other studies such as the following, it wasfound that adolescent girls have higher choles-terol levels than adolescent males:
●
●
●
On
Among 1,200 healthy males and females ofall ages in New York City, females aged13-17 years had higher mean cholesterollevels than males of comparable ages.zG
In a longitudinal study of 152 girls and 169boys in Utah, girls were found to havehigher mean cholesterol levels than theirmale counterparts.z T
Analysis by single years of age of data for885 volunteer and nonvolunteer male andfemale students aged 12-18 attendingpublic and parochial schools in Burlington,Vt., revealed that 12-year-old males hadhigher mean cholesterol levels than 12-year-old females; 13-year-old males and femaleshad approximately equal cholesterol levels;but beginning with age 14, however,females had higher levels than their malecounterparts.z 8
the other hand, a study conducted in ruralGeorgia showed no consistent differences in
8
mean cholesterol levels between male and femaleadolescents.z g
The decrease with age in mean cholesterollevels found in HES adolescent males was ob-served through age 16 both in the Burlington,Vt., studyz8 and in the survey of 613 schoolboys 11-18 years of age conducted in Sydney,AustraliansO Different patterns were observed,however, among adolescent males in two ruralsurveys: the one already mentioned above con-ducted in Geor~iaz g and another one conductedin Busselton, Australians1 The sharp increase incholesterol levels from age 16 to age 17 observedin U.S. females was also apparent among femalesin the Burlington, Vt., survey.zs The validity ofthe age-sex patterns found in Cycle III is furtherstrengthened by the fact that when mean choles-terol levels are plotted by age and sex for whiteand Negro youths separately, the curves forwhite youths are parallel to those for Negroyouths (figure 4), even though the curves for thelatter group are based on relatively small samplesizes in each age group.
The estimates of cholesterol levels for U.S.adolescents obtained from Cycle III of theHealth Examination Survey can be compared byextrapolation with those that have been re-ported for adults in HES Cycle 1.sz Table 13shows mean levels of cholesterol obtained foradults in Cycle I by age and sex along withstandard deviations of the population distribu-tion, and figure 6 shows these mean cholesterollevels in conjunction with those found for U.S.adolescents in Cycle III. In comparing the meanlevels of 17-year-old males (mean age 17.5years), the oldest age group of Cycle 111, withthose of 18- to 24-year-old males (mean age 21.5years), the youngest age group of Cycle I, wenote an average increase of 1.82 mg/ 100 ml peryear (figure 6). Similar examination of Cycle Idata for males in other age groups shows anaverage yearly increase of 3.27 mg/ 100 ml fromage 21.5 to age 30, 2.09 mg/100 ml from age 30to age 40, and much smaller increases thereafter.Thus, the mean cholesterol level obtained for17-year-old males in HES Cycle III is consistentwith the Cycle I finding that cholesterol levelsrise sharply with age in young men. Similaranalysis of the mean cholesterol level in 17-year-old females (mean age 17.5 years) with those of
— Male
---- Female
160 -
150-
Cycle III ~ Cycle I uo
1 , I 1
10 20 30 40 50 60 70 80
AGE IN YEARS
18- to 24-year-old Cycle I women (mean age21.5) shows practically no difference in meancholesterol levels between the two age groups(figure 6). Thus, the sharp increase in meancholesterol level observed in these HES databetween 16- and 17-year-old femalesd is notfollowed by continued increase during the ageinterval 18-24 years, and the gradual rise incholesterol level among women apparently doesnot begin until the mid or late twenties, whereasin males the rise is not only steeper but it beginsearlier.
‘The 80.2 percent response rate in 17-year-old females was
the lowest obtained in the survey (see appendix); and this mayaccount for some of the unexplainable findings observed in thisage-sex group.
Figure 6. Meen cholesterol levels for adults in HES Cycle I (1960-62) and youths in HES Cycle III (1966-70): United States.
Not only are the mean cholesterol levelshigher in U.S. adults than in U.S. adolescentsbut also the distributions have greater variabilityin adults than in youths 12-17 years of age(tables 3 and 15). The coefficient of variation ofthe cholesterol distribution averages about 17percent in adolescents, whereas it averages about22 percent in adults.
All other major studies of cholesterol inadults have also obtained mean levels that wereconsiderably higher than those obtained in thisCycle III survey of male and female youths12-17 years of age.zl *Zs~ze
Race
One of the most striking findings of thissurvey was that both male and female Negro
9
youths showed higher cholesterol levels thanwhite youths of comparable age and sex. Thesedifferences between white and Negro youthswere greater in males than in females and weregreatest at the upper percentiles of the distribu-tion. The average difference over all six agegroups in mean cholesterol levels between whiteand Negro youths was 7.8 mg/ 100 ml in malesand 2.7 mg/100 ml in females. Racial differencesalso occurred in all socioeconomic classes asmeasured by family income or education ofparent. Two other studies of cholesterol in U.S.adolescents have reported findings of little or nointerracial differences in mean cholesterol.zg’ aaHowever, the fact that one of these studies wasin a very rural areazg and the other was basedon a very small sample from both groups33makes it difficult to relate the findings to HESresults.
In the HES survey of U.S. adults (Cycle I),white males and females had consistently highermean cholesterol levels than Negro adults ofcomparable age and sex. Other studies haveshown racial differences in cholesterol levels. Astudy of cholesterol levels in SouthwesternAmerican Indians and white controls showedlower cholesterol levels among the Indians.34Outside of the United States, a survey of chil-dren and adolescents in South Africa amongwhite, Bantu, and Cape Coloured populationsshowed findings of higher cholesterol levels inthe white children than in either the Bantu orCape Coloured children.35
Annual Family Income
It appears from these data on U.S. adolescentsthat a positive relationship exists between meancholesterol levels and family income althoughthe relationship is not consistent for all age-sexgroups. However, when all age groups are com-bined (table 12), there is a consistent increase inthe mean cholesterol levels of both male andfemale youths with increase in family income.The group of males whose families had annualincomes of less than $3,000 had a mean choles-terol level of 170.6 mg/ 100 ml, whereas maleswhose families had annual incomes of $10,000or more had a mean level of 174.0 mg/100 ml,
reflecting an increase of 3.4 mg/ 100 ml from thelowest to the highest of family income. Thecomparable increase for females, was 6.8mg/100 ml, from a low of 172.7 mg/100 mlamong those whose families had annual incomesof less than $3,000 to a high of 179.5 amongthose whose families had incomes of $10,000 ormore.
Since 34.8 percent of the Negro youths asopposed to only 9.3 percent of the white youthscome from families with annual incomes of lessthan $3,000, the income group showing thelowest cholesterol levels, and since Negro youthshave higher mean cholesterol levels than whiteyouths at all income levels (table 12), the rela-tionships observed above between annual familyincome and mean cholesterol levels in the totalpopulation are greatIy influenced by the dis-proportionate number of Negro youths in thelower income groups. These two factors workingin opposite directions obscured and dampenedthe effective relationship between family incomeand cholesterol levels. Thus, we see that the’relationship between mean cholesterol andannual family income is stronger when examinedseparately in the white population than whenexamined in the total population (table 12). Inwhite males the gradient in mean cholesterollevel from the lowest to the highest incomegroup was 7.9 mg/100 ml, whereas in the totalmale population the gradient was only 3.4mg/ 100 ml. The comparable gradients in femaleswere 7.4 mg/ 100 ml for white females asopposed to 6.8 mg/100 ml for the total femalepopulation.
Other Studies
Several studies involving children and adoles-cents have investigated the relative influences ofheredity and environment as determinants ofcholesterol levels. The Busselton survey,~ 1 theTecumseh study,z 5 and a survey of residents oftwo Greek villagessG indicated considerableevidence of familial aggregation of cholesterollevels. In each of these studies, correlationalmethods were used to demonstrate parent-cKlldand sibling-sibling relationships. The general con-clusion of these studies was that the findings
10
were more compatible with a quantitive multi- in female pairs but not in male pairs.37 For thisfactorial mechanism of inheritance than with a particular group of twins, it was concluded that
single-gene genetic model. genetic determinants of cholesterol levels wereOn the other hand, a study of 108 like-sex generally of a smaller order of magnitude than
twin pairs showed that variation in cholesterol environmental determinants. Thus, it appears
level between monozygotic twins was signifi- that both heredity and environment influence
cantly smaller than that between dizygotic twins cholesterol levels.,4
000
REFERENCES
1National Center for Health Statistics: Origin, pro-gram, and operation of the U.S. National Health Survey.Vital and Health Statistics. PHS Pub. No. 1000-Seriesl-No. 1. Public Health Service. Washington. U.S. Gover-nmentPrinting Office, Apr. 1965.
2Nation~ Center for He~th Statistics: pl~ ~d
initial program of the Health Examination Survey. Vitaland Health Statistics. PHS Pub. No. 1000-Series l-No. 4.Public Health Service. Washington. U.S. GovernmentPrinting Office, July 1965.
3National Center for Health Statistics: Plan, opera-tion, and response results of a program of children’sexaminations. Vital and Health Statistics. PHS Pub. No.1000-Series l-No. 5. Public Health Service. Washington.U.S. Government Printing Office, Oct. 1967.
4Nation~ Center for Health Statistics: Plan ~doperation of a Health Examination Survey of U.S.youths 12-17 years of age. Vital and Health Statistics.PHS Pub. No. 1000-Senes l-No. 8. Public Health Ser-vice. Washington. U.S. Government Printing Office,Sept. 1969.
5National Center for Health Statistics: Height andweight of youths 12-17 years, United States. Vital andHealth Statistics, Series 1l-No, 124. DHEW Pub. No.(HSM) 73-1606. Health Services and Mental HealthAdministration. Washington. U.S. Government PrintingOffice, Jan. 1973.
6National Center for Health Statistics: Body weight,stature, and sitting height: White and Negro youths12-17 years, United States. Vital and Health Statistics.Series 1l-No. 126. DHEW Pub. No. (HRA) 74-1608.Health Resources Administration. Washington. U.S. Gov-ernment Printing Office, Aug. 1973.
7Nation~ Center for Heal~ Statistics: Skinfoldthickness of youths 12-17 years, United States. Vital and
Health Statistics. Series 1l-No. 132. DHEW Pub. No.(HRA) 74-1614. Health Resources Administration.Washington. U.S. Government Printing Office, 1974.
8Nation~ Center for Health Statistics: Hematocntvalues of youths 12-17 years, United States. Vital andHealth Statistics. Series 1 l-No. 146. DHEW Pub. No.(HRA) 75-1628. Health Resources Administration.Washington. U.S. Government Printing Office, Dec.1974. ,
9National Center for Health Statistics: Serum uricacid values of youths 12-17 years, United States. Vitaland Health Statz”stz”cs. Series 1l-No. 152. DHEW Pub.No. (HRA) 76-1634. Health Resources Administration.Washington. U.S. Government Printing Office, Aug.1975.
10Herskowits, M.J.: The Amen-can Negro; a Study inRacial Crossing. New York. Alfred A. Knopf, 1928.
11 GIxs, B., ~d Li, C.C.: The dynamics of raci?dintermixture-an analysis based on the American Negro.Am.J.Hum. Genet. 5(1):1-20, Mar. 1953.
12 Roberts, D. F.: The dynamics of racial intermixturein the American Negro: Some anthropological consider-ations. Am. J.Hum. Genet. 7(4):36 1-367, Dec. 1955.
13 Roberts, D. F., and Hiorns, R. W.: The dynamics ofracial intermixture. Am.J.Hum. Genet. 14(3) :261-277,Sept. 1962.
14 Reed, T. E.: Caucasian genes in American Negroes.Science 165(3895): 762-768, Aug. 22, 1969.
15 Eavenson, E., Grier, O.T., Cisson, J.G. and Witter,R. F.: A semiautomated procedure for the determinationof serum cholesterol using the Abell-Kendall method. J.Am. Oil Chem. Sot. 43(12):652-656, 1966.
16Henry, Richard J.: Clinical Chemistry, Principlesand Technics. New York. Hoeber Medical Division,Harper & Row, 1965.
11
17 Fredrickson, D. S., and Levy, R.L: “Fa&lial hyper-Iipoproteinemia,” in J.B. Stanbury, J.B. Wyngaarden,and D.S. Fredrickson, eds., The Metabolic Basis ofInherited Disease, 3rd ed. New York. McGraw-Hill BookCo., 1972. pp. 545-614.
18Sherwin, R.: The epidemiology of atherosclerosisand coronary heart disease. Postgrad. Med. 56(6):81-87,1974.
19 Smith, E.B.: lntimd and medial lipids in hum~aortas. Lancet, Vol. 1, Part II: 799-803, Apr.-June,1960.
Zo%nith, E.B.: The influence of age and atheroscle-rosis on the chemistry of aortic intima. Part I. TheLipids. J. Atheroscler. Res. 5:224-240, 1965.
21 K-el, W-B+, ad Gordon, “rQ,eds.: Me~s at each
exam ination and inter-examination variations ofspecified characteristics. Section 29, The FraminghamStudy, exam l-exam. 10. DHEW Pub. No. (NIH) 74-478,Dec. 1973.
22 Cdson, L. A., and B6ttiger, L. E.: Ischaemic heart-disease in relation to fasting values of plasma triglyc-erides and cholesterol: Stockholm Prospective Study.Lancet, Vol. 1, Part II: 865-868, Apr.-June, 1972.
2~HoIman, R. L., et al.: The natural history of athero-sclerosis: The early aortic lesions as seen in New Orleansin the middle of the 20th century. Am. J.Path.34(2):209-236, Mar. 1958.
24 Enos, W. F., Holmes, R. H., and Beyer, J.: Coronarydisease among United States soldiers killed in action inKorea. Preliminary Report. J.A.M.A. 152(12):1090-1093, July 1953.
25Johnson, B.C., Epstein, F.H., and Kjelsberg, M.O.:Distributions and familial studies of blood pressure andserum cholesterol levels in a total community—Tecumseh, Mlchlgan. ]. Chron.Dis. 18, 147-160, Feb.1965.
26Adlersberg, D.A., Schaefer, L. E., SteinbeW, A. G.,and Wang, C.-I.: Age, sex, serum lipids, and coronaryatherosclerosis. J. A.M.A. 162(7):619-622, Oct. 1956.
27M~lig~, C.A., Wilcox, E.B., and Galloway, L. S.:Serum cholesterol and physical characteristics of pre-adolescents and adolescents. J. Am. Diet. Assoc. 49(4):309-315, Oct. 1966.
28Clarke, R. P., et al.: Interrelationships betweenplasma lipids, physical measurements, and body fatness
of adolescents in Burlington, Vermont. Am.J. Clin.Nutr.
J23(6 :754-763, June 1970.
2 Hames, C.G., and Greenberg, B.G.: A comparativestudy of serum cholesterol levels in school children andtheir possible relation to atherogenesis. Am.J. PublicHealth 51(3):374-385, Mar. 1961.
SoHickie, J. B<, et al.: Serum cholesterol and serumtriglyceride levels in Australian adolescent males. Med. J.Aust. 1,825-828, 1974.
31 Godfrey, R. C., Stenhouse, N.S., CuUen, K.J., andBlackman, V.: Cholesterol and the child: Studies of thecholesterol levels of Busselton school children and theirparents. Aust.Paediat.J. 8, 72-78, 1972.
32 Nation~ Center for Health Statistics: Serum ch~Iesterol levels of adults, United States– 1960-1962. Vitaland Health Statistics. PHS Pub. No. 1000-Series 1l-No.22. Public Health Service. Washington. U.S. GovernmentPrinting Office, Mar. 1967.
38Baker, H., et al.: vitm~s, total cholesterol, andtriglycerides in 642 New York City school children.Am.J. Clin.Nutr. 20(8):850-857, Aug. 1967.
34 Sievers, M-L.: Semm cholesterol levels in south-western American Indians. J. Chronic Dis. 21(2): 107-115,1968.
“35Du Plessis, J. P., Vivier, F.S., and De Lange, D.J.:The biochemical evaluation of the nutrition status ofurban school children aged 7-15 years: Serum choles-terol and phospholipid levels and serum and urinaryamylase activities. South Afn”can Medical Journal. 41,1216-22, 1967.
36Mayo, O., Fraser, G. R., and Stamatoyannopoulos,G.: Genetic influences on serum cholesterol in twoGreek Villages. Hum. HerecL 19, 86-99, 1969.
37 ~d, B.M., et al.: Study of serum lipid levek intwins. Cardiovasc.Res. 2, 148-156, 1968.
38Nation~ Center for He~th Statistics: s~pledesign and estimation procedures for a national healthexamination survey of children. Vital and Health Statis-tics. Series 2-No. 43. DHEW Pub. No. (HRA) 74-1005.Health Resources Administration. Washington. U.S. Gov-ernment Printing Office, Aug. 1971.
89 National Center for Health Statistics: Replication:An approach to the analysis of data from complexsurveys. Vital and Health Statktics. PHS Pub. No.1000-Series 2-No. 14. Public Health Service. Washington.U.S. Government Printing Office, Apr. 1966.
12
LIST OF DETAILED TABLES
,
Table 1, Estimated number of youths aged 12-17 years in the population, by cholesterol group, sex, and age: United
States, 1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2. Percent distribution of youths aged 12-17 years by cholesterol group, according tosexand age: United States,1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3, Unweighed and weighted sample sizes, mean cholesterol, standard deviation, standard error, and selected
percentiles,bysexandage: United States,1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4. Mean cholesterol and standard error of the mean for youths aged 12-17 years, by geographic region, sex, andage: UnitedStates,1966-70. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5. Estimated number of white youths aged 12-17 years in the population by cholesterol group, sex, and age:UnitedStates, 1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6. Estimated number of Negro youths aged 12-17 years in the population, by cholesterol group, sex, and age:UnitedStates, 1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7. Parcent distribution of white youths aged 12-17 years by cholesterol group, according tosexandaga: UnitedStates, 1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8. Percent distribution of Negro youths aged 12-17 yaars by cholesterol group, according tosexandaga: UnitadStates, 1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9. Unweighed and weighted sample sizes, meen cholesterol, standard deviation, standard error, and selectadpercentiles, byrace, sex, andage: United Statas, l966-7O . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10. Weightad sample size, maan cholesterol, andstandard error of themean foryouths agad.12-17 yaars, by age,sex,andannualfamilyincoma: UnitedStatas,1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11. Weighted sample size, maan cholesterol, and standard arrorof the mean for youths agad 12-17 years, by aga,sex,andeducationofparant: UnitadStates,1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12. Weighted sampla size, mean cholesterol, and standard error of the mean for youths aged 12-17 years, by race,
sex, annual family income, and education of parent: United States, 1966-70 . . . . . . . . . . . . . . . . . .
13. Weighted sample size, mean cholesterol, andstandard error of the mean foryouths agad 12-17 years, by age,race, sex, andannual family incoma: United States, 1966-70 . . . . . . . . . . . . . . . . . . . . . . . . . .
14. Waighted sample size, mean cholesterol, and standard error of the mean for youths aged 12-17 years, by age,race, sax, andeducation ofparent: United States, l966-7O . . . . . . . . . . . . . . . . . . . . . . . . . . .
‘ 15, Mean cholesterol levels andstandard deviations of thepopulation distribution foradults, bysexandage: UnitadStates, 1960-62 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Page
14
14
15
15
16
17
18
19
20
21
21
22
23
24
25
●
13
Tablel. Estimated number ofyouths agd12-17years inthepopulation, bycholesterol group, sex, andage: United States,l966-7O
Male I Female
Cholesterol group12
years
13 I 14
years years
15 16 17 12 13 14 15 16 17
years years years years years yaars years years yaars
Estimatad number of youths in population in thousends
1,951 1,900 1,836 1,764 1,970
7
10
172083
131201
317299279226141
8185301817
34
1,946 1,901 1,8612,032 2,006 1,769 1,746All groups . . . . . .
Under 100mgpercent . .
100-108 mgpercent . . . .
110-l 19mgpwcent . . . .120-129 mgpercent . . . .130-138 mg percent . . . .140-149 mg ~rcent . . . .150-159 mg percent . . . .
160-169 mg ~rcent . . . .170-179 mg percent . . . .180-189 mg percent . . . .190-199 mg percent . . . .200-209 mg p?rcent . . . .210-219 mg percent . . . .220-229 mg percent . . . .230-239 mg percent . . . .240-249 mg percent . . . .250-259 mg percant . . . .260-269 mg psrcent . . . .270 mg percent and over .
-83
2465
116224251
255273217184
94120
70372921
510
6194759
143228243
289227226179102
7844271611
36
4
164557
171210271
290230229130
7770233825
3
94
6
14
5171
139210252
276250229137
5664262113
787
1115
3268
113194232
242221186160
92983722
767
10
5
33775
104142222
227257278184142
88602417
81613
611
1419
105181222
218287238197116101
52212310
522
4
103128
115204
222
273214156144115
514252301233
3
174386
136179
269291287220183113
85522511
527
155397
185240
245246238170159126
784436
653
183783
173232
179280232187
8798826110
7
915
Table 2. Percent distribution of youths aged 12-17 years by cholesterol group, according to sex and age: United States, 1966-70
Male Female
12 13 14 15 16 17 12 13 14 15 16 17
years years years years years years yeers years years yeers years years
Cholesterol group
Percent distribution
All groups . . . . . . . . . 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 I 100.0 100.0 100.0
0.3 0.30.2 0.62.0 0.84.0 1.06.5 5.77.4 9.8
11.7 12.0
11.9 11.813.5 15.514.6 12.9
9.7 10.67.4 6.34.6 5.53.1 2.8
1.3 1.10.9 1.30.4 0.50.9 0.3
0.7 1.2
Under 100 mg,percant . .100-109 mg percant . . . .110-119 mgpwcent . . . .120-129 mg percent . . . .130-139 mg percent . . . .140-149 mg percant . . . .150-159 mg percant . . . .
160-169 mg percent . . . .170-179 mg percent . . . .180-189 mg pwcent . . . .190-199 mg pwcant . . . .200-209 mg percent . . . .210-219 mg percent . . . .
220-229 mg percent . . . .230-239 mg percent . . . .240-249 mg percent . . . .250-259 mg p?rcent . . . .260-269 mg percent . . . .270 mg ~rcent and over .
0.40.11.23.25.8
11.212.5
12.7
13.610.8
9.24.76.0
3.51.91.4
1.00.2
0.5
0.31.02.43.07.3
11.712.4
14.8
11.611.69.25.24.0
2.21.40.80.60.20.3
0,20.92.43.09.0
11.114.315.3
12.112.1
6.84.13.7
1.2
1.91.30.10.50.2
0.30.82.83.97.6
11.413.7
15.0
13.612.5
7.43.13.51.4
1.20.70.40.40.4
0.60.81.83.96.4
11.013.2
13.7
12.511.2
9.15.26.62.1
1.20.40.30.40.5
0.30.50.91.04.26.6
10.2
16.1
15.214.211.5
7.14.1
4.31.50.90.90.20.2
0.2
0.61.81.66.6
11.7
12.715.612.2
9.68.36.6
2.9
2.43.01.70.7
1.9
0.20.82.14.26.78.8
13.2
14.314.1
0.72.75.09.5
12.3
12.612.712.1
8.78.16.54.0
2.31.90.30.30.2
1.02.14.69.7
12.9
10.0
15.613.010.4
4.95,5
4.63.40.60.40.5
0.8
10.89.05.54.22.51.20.50.31.3
NOTE: Percents may not add to 100.0 due to rounding.
Tabla 3. Unweighed and weighted semple sizes, mean cholesterol, stendard deviation, standard error, and salacted percentiles, by sex
and aw: United Statas, 1966-70
Percentilen N E Sx Sk
5 10 25 50 75 90 95Sex and age
Male
12 years . . . . . . . . . . . . .13years . . . . . . . . . . . . .14 years . . . . . . . . . . . . .15 years . . . . . . . . . . . . .16 years . . . . . . . . . . . . .17 years . . . . . . . . . . . . .
Female
12 yaars . . . . . . . . . . . . .13 years . . . . . . . . . . . . .14 yaars . . . . . . . . . . ... .16 years . . . . . . . . . . . . .16 years . . . . . . . . . . . . .17 years . . . . . . . . . . . . .
Cholesterol in mg/100 ml
843626618613556489
547582586503536469
2,0322,0061,9511,9001,8361,764
1,9701,9461,9011,8511,7891,746
181.3174.7169.9188.516,7.7170.8
178.0177.7175.3176.8178.1184.5
29.8531.1930.2029.8428.57
30.36
27.8729.4631.8630.2630.0134.16
1.401.171.561.401.761.70
0.981.741.591.591.861.31
137.4128.3125.1126.3123.3125.3
135.8133.4125.8133.8136.1141.0
146.4138.4134.7135.4134.7133.7
143.8141.5137.1141.5143.3148.4
160.9152.8149.4148.5149.1150.4
161.0156.0155.1156.1156.3
161.8
?79.5171.8167.8165.5165.8169.1
176.2174.7173.8175.3175.7179.2
199.3193.2188.8185.4184.0189.8
194.1198.5192.5194.3196.2201.8
219.4217.3207.7207.6201.6211.6
214.2218.8212.8213.4221.0
226.6
232.5230.5221.4220.5217.6220.4
225.6229.6226.1225.6231.724a5
NOTE: n=sample size; N= estimated number of youths inthou~nds; ~=mean; sx=mandard deviation; s~=standard errorofthe mean.
Table 4. Meen cholesterol and standard error of themean for youths aged 12-17 years, bygeographic region, sex, and age: United
States, 1966-70
9
Sex and age
Male
12 years . . . . . . . . . . . . . . . . . . . . . . . . .13 yaars . . . . . . . . . . . . . . . . . . . . . . . . .14 years . . . . . . . . . . . . . . . . . . . . . . . . .15 years . ...,.,. . . . . . . . . . . . . . . . . .
16 yaars . . . . . . . . ,., . . . . . . . . . . . . . .17 years . . . . . . . . . . . . . . . . . . . . . . . . .
Female
12 years . . . . . . . . . . . . . . . . . . . . . . . . .13 years . . . . . . . . . . . . . . . . . . . . . . . . .14 years . . . . . , . , . . . . . . . . . . . . . . . . .15 yaars . . . . . . . . . . . . . . . . . . . . . . . . .
16 years . . . . . . . . . . . . . . . . . . . . . . . . .17 years . . . . . . . . . . . . . . . . . . . . . . . . .
Geographic region
TotalNorth- Mid-
South Wasteast west
Mean cholesterol in mg/100 ml
Wast
Standard error
181.3
174.7
169.9
168.5
167.7
170.8
178.0
177.7
175.3
176.8
178.1184.5
185’.2175.2172.6169.9
172,6
175.6
176.5183.9179.5180.0184.7184.6
181.1
173.8
174.3
169.2
165,3
165.4
180.8
177.2
170.9
177.6
174.5184.4
182.4
175.3
166.3
168.0
163.9167.6
175.7173.1176.2
179.6179.3185.2
177.0
174.5
165.9
166.9
170.4
175.3
177.9
177.1
175.6
170.7
175.9184.0
1.401.171.561.40
1.761.70
0.981.741.591.591.861.31
2.362.693.792.25
2.253,42
2.022.493.01
3.262.842.58
1.901.982.923.50
2.76
1.91
2.213.693.02
2.483.651.87
3.253.102.153.06
4.84
3.28
3.843.673.09
3.323.973.64
4.121.582.163.76
3.222.61
1.523.873.034.133.573.50
16
Table 5. Estimated number of whita youths aged 12-17 yaarsin the population by cholesterol group, sex, andaga: United States,
Cholesterol group
Al I groups . . . . . .
Under 100mg percent . .
100-109 mgpercent . . . .
110-l 19mgpercant . . . .
120-129 mgpxcent . . . .
130.138 mgpercent ,.. .
140-149 mg percant . . . .
150-159 mg percent . . . .
160-169 mg percent . . . .
170-179 mg percent . . . .
180-189 mg ~rcent . . . .190-199 mg percent . . . .
200-209 mg percent . . . .
210-219 mg percant. . . . .
220-229 mg percent . . . .
230-239 mg pwxmt . . . .240-249 mg percent . . . .
250-259 mg percent . . . .
260-269 mg percent . . . .
270 mg percent and over
1966-70
Male I Female
12
yaars
1,747
3
17
35
81
118
161
212
266
253
193
160
92
63
4221
8
3
19
13
years
1,729
8
20
55
102
211
201
230
241
205160
69
104
39
3422
11
5
10
14
yaars
15 I 16
years years
17 I 12
years years
13 14 15 16 17
years years years years years
Estimated number of youths in population in thousands
1,686
6
17
43
56
125
208
216
258
190
166
160
88
69
28
2713
8
3
3
1,646
4
14
45
49
157
195
257
237
178
200
107
58
54
23
31
25
3
9
3
1,594
6
14
37
64
132
201
233
225
218
186
110
47
57
18
18
97
8
7
1,528
1115
32
54
101
168
195
232
190
178
133
73
74
32
197
6
4
3
1,685
7
10
14
20
70
102
176
270
246
242
206
123
68
68
2812
14
3
4
1,667
12
42
92
162
212
214
223
213
131
129
98
62
3535
3
3
1,633
2
3
27
66
98
118
204
180
221
245
160
121
73
51
2112
3
13
13
1,594
~
6
11
8
19
85
167
199
177
251
203
180
90
93
40
1419
10
5
19
1,542
18
23
78
159
197
157
250
202148
67
91
68
516
5
7
15
1,502
4
8
28
23
88
173
187
263
199
129
132
90
42
28
36
27
12
30
16
Tabla 6. Estimated number of Negro youths aged 12-17 years in the population by cholesterol group, sex, and age: United States,
Cholesterol group
All groups . . . . . . . . . .
Under 100mg percent . . . . . .
100-109 mg percent . . . . . . .
110.119 mg percent ,., ,.. .
120-129 mg percent . . . . . . .
130-138 mg percent . . . . . . .
140-149 mg percent . . . . . . .
150-159 mg percent . . . . . . .
160-169 mg percent . . . . . . .
170-179 mg percent . . . . . . .
180-189 mg percent . . . . . . .
190-199 mg percent . . . . . . .
200-209 mg percent . . . . . . .
210-219 mg percent . . . . . . .
220-229 mg percent . . . . . . .
230-238 mg percent . . . . . . .
240-249 mg parcent . . . . . . .
250-259 mg percent . . . . . . .
260-269 mg percent . . . . . . .
270 mg parcant and over . . . .
1966-70
Male Female
12 13 14 15 16 17 12 13 14 15 16 17
years years years years years years years years years years years years
Estimated number of youths in population in thousands
280~
8
5
18
18
57
24
34
27
19
21
22
103
3
3
8
262
33
1014133824291224251631
37
10
256
2
4
3
18
20273134521914
815
33
2
241 231
2 -
149 7
14 416 914 1950 5052 3323 3923 2719 1013 8
85 -
3
2 -
225
14122637103117241924
3
26
-
272
3
13292547533718121312
223
275
275
232832232339302716
9335
266
3
1010
52418473630232115
83553
I
235 243
6 -14
21 512 1418 3531 2237 3037 2717 3723 19
8 712 13
7 104 5
22
3 -
237
2
3
6
26
31
31
10
15
36
12
25
9
14
10
3
2
17
Table 7. Percent distribution Of white youths agad 12-17 years by cholesterol group, according to sex and age: United States, 1966-70
Cholesterol group
All groups . . . . . .
Under 100 mg percent . .
100-109 mg percent . . . .
110-l 19mgpercent . . . .
120-129 mg percent . . . .
130-139 mg percent . . . .
140-149 mg percent . . . .
150-159 mg percent . . . .
160-169 mg percent . . . .
170-179 mg percent . . . .
180-189 mg percent . . . .
190-199 mg percent . . . .
200-209 mg percant . . . .210-219 mg percent . . . .
220-229 mg parcant . . . .
230-239 mg percent . . . .240-249 mg percent . . . .
250-259 mg percent . . . .
260-269 mg percent . . . .
270 mg percent and over .
Male Female
12 13 14 15 16 17 12 13 14 15 16 17years years years years years years years years years years years years
Percent distribution
100.0
0.2
1.0
2.0
4.6
6.8
9.2
12.2
15.3
14.5
11.1
9.1
5.3
3.6
2.4
1.2
0.4
0.1
1.1
100.0
0.5
1.2
3.2
5.9
12.2
11.6
13.3
13.9
11.9
9.2
4.0
6.0
2.2
2.0
1.3
0.6
0.3
0.6
100.0
0.3
1.0
2.5
3.3
7.4
12.3
12.8
15.3
11.3
10.0
9.5
5.2
4.1
1.7
1.6
0.8
0.5
0.2
0.2
100.0
0.3
0.9
2.7
3.0
9.5
11.8
15.6
14.4
10.8
12.2
6.5
3.53.3
1.4
1.9
1.5
0.2
0.5
0.2
100.0
0.4
0.9
2.3
4.0
8.3
12.6
14.6
14.1
13.7
11.7
6.9
2.9
3.6
1.1
1.1
0.6
0.4
0.5
0.4
100.0
0.7
1.0
2.1
3.6
6.6
11.0
12.8
15.2
12.4
11.7
8.7
4.8
4.9
2.1
1.2
0.5
0.4
0.3
0.2
100.0
0.4
0.6
0.8
1.2
4.2
6.1
10.4
16.0
14.6
14.4
12.3
7.34.1
4.0
1.7
0.7
0.6
0.2
0.2
100.0
0.7
2.5
5.5
9.7
12.7
12.8
13.4
12.8
7.9
7.7
5.9
3.7
2.1
2.1
0.2
0.2
100.0
0.10.2
1.7
4.0
6.0
7.2
12.5
11.0
13.5
15.0
9.8
7,4
4.5
3.1
1.3
0.7
0.2
0.8
0,8
100.0
0.4
0.7
0.5
1.2
5.3
10.4
12.5
11.1
15.7
12.7
11.3
5.6
5.8
2.5
0.9
1,2
0.6
0.3
1.2
100.0
1.2
1.5
5.0
10.3
12.8
10.2
16.2
13.1
9.6
4.4
5.9
4.4
3.3
0.4
0.3
0.5
1.0
100.0
0,3
0,s
1.9
1,5
5,9
11,5
12.4
17.5
13.2
8.6
8.8
6.0
2.8
1.9
2.6
1.8
0,8
2.0
NOTE: Percents may not add to 100.0 due to rounding.
18
Table 8. Percent distribution of Neriro youths aged 12-17 years by cholesterol group, according to sex and age: Unitad States, 1966-70
Cholesterol group
All groups . . . . . .
Under 100 mg percent . .
100-109 mg percent . . . .
110-l 19mgpercent ., . .
120-129 mg percent . . . .
130-139 mg percent . . . .
140-149 mg percent . . . .
150-159 mg percent . . . .
160-169 mg percent . . . .
170.179 mg percent . . . .
180-189 mg percent . . . .
190-199 mg percent . . . .
200-209 mg percent . . . .
210-219 mg percent . . . .
220-229 mg percent . . . .
230-239 mg percent . . . .
240.249 mg percent . . . .
250.259 mg percent . . . .
260-269 mg percent . . . .
270 mg percent and over .
Male Female
12 13 14 15 16 17 12 13 14 15 16 17
years years years years years years years yaars years years yaars years
Percent distribution
100.0
2.8
1.8
6.6
6.5
20.2
8.7
12.2
9.5
6.6
7.4
7.9
3.6
1.2
1.1
1.0
2.8
100.0
1.01.3
3.7
5.2
4.8
14.7
9.2
11.0
4.6
9.3
9.4
6.2
12.0
1.2
2.6
3.8
100.0
0.9
1.5
1.3
7.0
7.7
10.6
12.1
13.3
20.4
7.4
5.4
3.2
5.9
1.1
1.1
0.9
100.0
0.9
3.6
5.8
6.5
5.6
20.7
21.5
9.6
9.7
8.0
5.2
2.1
0.8
100.0
6.0
3.1
1.7
3.9
8.3
21.8
14.1
16.8
11.5
4.2
3.4
3.7
1.4
100.0
6.2
5.5
11.3
16.4
4.4
13.7
7.7
10.7
8.3
10.5
1.3
1.1
2.8
I00.0
1.1
4.9
10.5
9.2
17.4
19.7
13.6
6.8
4.4
4.8
4.6
0.9
0.9
1.3
100.0
0.8
2.6
2.0
8.4
10.3
11.4
8.4
8.3
14.0
10.9
9.9
5.8
3.2
0.9
1.1
1.8
100.0
1.1
3.6
3.7
2.0
8.9
6.7
17.6
13.5
11.3
8.8
7.9
5.7
3.1
1.0
1.9
2.01.1
100.0
2.6
8.7
4.9
7.7
13.3
15.7
15.7
7.1
9.7
3.6
5.1
2.9
1.7
1.3
100.0
5.9
2.1
5.7
14.4
9.0
12.5
11.2
15.2
8.0
2.8
5.4
4.2
1.9
0.71.0
100.0
1.01.12.3
11.2
13.3
13.0
4.2
6.3
15.4
5.0
10.6
4.0
5.9
4.4
1.3
1.0
NOTE: Percents may not add to 100.0 due to rounding.
19
Table 9. Unweighed and weighted sample sizes, mean cholesterol, standard deviation, standard error, and salected percantilas, by race,
Race, sax, and age
Whita male
12 years . . . . . . . . . . . . .13 years . . . . . . . . . . . . .14 years . . . . . . . . . . . . .15 years . . . . . . . . . . . . .16years . . . . . . . . . . . . .17 years . . . . . . . . . . . . .
White famale
12 years . . . . . . . . . . . . .13years . . . . . . . . . . . . .14 years . . . . . . . . . . . . .15years . . . . . . . . . . . . .16yaars . . . . . . . . . . . . .17 years . . . . . . . . . . . . .
Nemo male
12years . . . . . . . . . . . . .13years . . . . . . . . . . . . .14years . . . . . . . . . . . . .15years . . . . . . . . . . . . .
16years . . . . . . . . . . . . .17years . . . . . . . . . . . . .
Negro female
12years . . . . . . . . . . . . .13years . . . . . . . . . . . . .
14years . . . . . . . . . . . . .15years . . . . . . . . . . . . .16years . . . . . . . . . . . . .17years . . . . . . . . . . . . .
sex, andaga: United States, 1966-70
Percentilen N z 5X SF
5 10 25 50 75 90 95
540542527525496417
455490484425441393
101808884
5769
8891
101739374
1,7471,7291,6861,6461,5941,528
1,6851,6671,6331,5941,5421,502
280262256241
231225
272275
266235243237
180.2173.5168.8167.3166.7169.7
178.0176.6175.0176.5177.7184.2
187.2183.1176,9175.6
173.3176.8
175.4185.5177.0179.7180.9185.9
29.2230.4029.5729.4628.5929.62
27.9728.9231.7530.2729.9133.97
33.0235.4133.6231.50
26.6433.94
25.9831.0832.6030.7130.7234.91
Cholesterol
1.591.261.461.481.751.40
1.231.821.671.582.021.23
3.013.804.743.04
4.466.44
3.873.03
2.534.252.464.27
137.3130.7124.1125.7124.0123.0
135.3133.6126.9134.4136.9
141.2
141.1126.8134.1134.3
118.8127.9
136.7136.4122.5131.7127.8140.4
)mg/100ml
146.0138.2133.6134.6134.7133.8
144.0141.2137.0141.5143.3149.3
146.8136.6137.7139.1138.5133.0
143.1145.2136.5136.0142.6146.2
160,3152.1148.4147.0147.8150.3
161.2155.4154.8155.7155.9162.7
163.0155.2158.3161.9
159.8150.2
158.9161.7158.0161.5158.5157.6
178,7171.3165.8163.3164.7168.4
176.6173.8173.5175.2175.0
178.7
183.4177.0174.8173.3
174.0176.2
172.7187.8
174.6178.2180.9184.1
197.3191.3188.2184.6183.5187.2
194.7196.8192.2193.7195.2201.1
208.1210.8190.6190.6
186.4196.4
187.7207.2
194.1199.6199.3215.9
216.1214.4206.4207.4201.0208.1
213.8217.6212.2213.0220.3224.3
229.4228.5217.7207.2
206.3217.8
213.7225.1215.6221.2224.0234.3
231.1228.9220.0221.6216,1219.6
225.6228.5224.4223.8231.5248,6
241.8242.2227.5218.6
218.2236.4
224.8235.4239.2231.9231.9245.6
NOTE: n =sample size; N= estimated number of youths inthousands; ~=mean; sY=standard deviation ;sp=standard error of
the maan.. . ,.
20
Tabla 10. WalQhtedmmplesize, m=ncholesterol.8 ndstan&rd error of themean foryouths Wd12-17vears, bve, sx, andannual family inmme:Unit& States,l*7O-
12 years 13 years 14 years 15 years 16 years3SXand annual family income
17 years
N x SF N F $2 N x % N J? SF N z $E N z SF
Mole—I
Cholesterol in IWI1OO ml
All Incomes . . . . . . . .
L*I: thm $3,0W . . . . . . . . .$3J200.34,999 . . . . . . . . . . .$5,01X7.$9,8s9 . . . . . . . . . . .$l0,0000rmcm . . . . . . . . .Don’t know . . . . . . . . . . . . .Blank orrafused . . . . . . . . . .
Female
Alllnc.mas . . . . . . . .
L**s than $3,000 . . . . . . . . .$3,cmo-$4,999 . . . . . . . . . . .$6,000.$B,999 . . . . . . . . . . .SI0,0000rmOre . . . . . . . . .Don’t know . . . . . . . . . . . . .Blank orref used . . . . . . . . . .
2,032—
180306669575
5545
181.3 1.40 I 2,005 174.7 I 1.17 1.9511 159.9 1.551.900 158.5 1.40 1.828 1.70167.7 I 1.76- . _ _ —f77.2179.3182.7180.5167.0186.4
5.57 212254767619
6580
176.4 5.63175.1 5.10173.6 2.10175.5 1.62176.3 4.11171.0 5.12
m24074659710067
163.8171.2169.8172.2163.4174.1
3.665..562.242.606.387.06
241219719598
7648
164.7169.4166.3172.7174.4153.5
5.032.991.692.083.596.56
201
255719516
6255 1
163.9 5.03163.7 4.40167.3 2.31170.7 2.92170.1 5.55173.4 6.92
17s.1 1.86
17s.9 3.4s172.6 3.22177.3 2.66179.5 2.92177.8 3.64188.5 11,50
226245692479
4974
178.2167.0169.2171.6163.0176.1
3.705.081.902.999.10B.14
3.561.95Z.oa5.13
‘1O.92
41,901 175.31,970-
2572747226KI
7937
178.0_
172.9177.5177.5160.2179.6183,9
=L=
T177.7 1.74
166.9 4.31179.6 2.71181.3 2.31176,22.68183.1 9.08176,2 11.03
1.59 1,851 176.8 1.59 1,7s9 1,746 134.5 1.31-. _ . . - -
3.84 25a4.33 3012.50 8022.00 4844.39 46
42,82 35
213 154.s3M 169.5720 160,9554 176.1
72 178.242 157.5
4.244.221.822.627.21
11.13—
2582647005336332
176.4173.2175.31s1.0171.3161.7
5.133.162.322.575.21
10.17
246156650594
6350
1S2206536628
92S2
176.7194.9182.91s3.0190.5197.5
6.157.042.432.295.619.69
NOTE: N = em!mated number of youths in thousands;~ = mea., SF = standarderror of the mean.
Tablo 11,Welght8dsamplesize,mean cholesterol,and standard error of the mean for youths aged 12-17 years,by age,tex, and education of parent United 8tates, 185$-70
Scx and education of parent
Cholesterol In mg/100 mlMde—
All aducation IJOUIIS .
Lessthan6ynars . . . . . . . . .8.11 War$ . . . . . . . . . . . . . . .12 years . . . . . . . . . . . . . . . .13 yearaor mora . . . . . . . . .Unknown . . . . . . . . . . . . . . .
Femalm
All education groups .
Lnssthan 8 years . . . . . . . . .E.ll yams. . . . . . . . . . . . . . .12v9mt . . . . . . . . . . . . . . . .13ymrs0rm0r0 . . . . . . . . .Unknown . . . . . . . . . . . . . . .
1,951 169.9 1.56 1,600 168.5
172 161.3 5.66 196 161.8474 172.2 3.27 487 167.3730 169.1 1.47 727 170.9 m2,032 181.3 1.40 2,00% 174.7 1.17
176 177.1 4.77 175 166.8 3.79543 182.3 2.89 502 176.6 3.24736 180.S 2.04 760 175.0 2.2s Tl_
1.40 1,636 167.7
4.10 217 164.12.70 474 167.01.96 630 168.6
56216
182.5 I 2.96 I 527 I 174.6 I 2.24 549 172.2 2.29 472 169.625 157.4 18.05 16 144.7
1,901 175.3 1.59 1,651 176.8
2.28 480 169.112.26 35 182.7
1.59 1,769 178.14+2.35 286 173.4 2.115.79 54 174.2 9.44
1.86 1.746 1S4.5 1.31
5.18 146 185.9 11.662.49 465 183.4 2.(332.54 525 164.7 2.624.17 542 161.9 2.67
25.53 67 209.1 17.72
186.2 45.40 23 182.1 9.27
1,970 178.0 I 0.98 I 1,946 I 177.7 I 1.74
m Tr6.16 167 175.22.80 453 173.22.66 623 177.01.32 612 163,8
211430729509
174.1 4.58 228 172.9 8.39176.1 2.82 507 177.4 1.70178.8 1.86 727 161.3 2.43180.2 1.84 463 175.3 2.45
41 I 176.5 1G.39 97 I .1 . 21 184.0 29.66 22 170.6 42.33 24 195.7.- 1 I
NOTE: N. tdtlmated number of youths in thousands; ~ = mean;s~ = standard error of the mean.
21
●
Table 12. Weighted sample size, mean cholesterol, and standard error of the mean for youths aged 12-17 years, by race, sex, annual
family income, and education of parent: United States, 1966-70
Variable
MALE
Annual family income
Lessthan $3,000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.$3,000-.$4,999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
$5,000-$9,999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
$10,000 ormore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Education of parent
Lessthan8 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8-11 years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13yearsor more . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FEMALE
Annual family income
Less than $3,000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
$3,000-$4,999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .$5,000-$9,999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
$l0,000ormore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Education of parent
Lessthan8years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8-n years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
12years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13 years or more . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Total White I Negro
Cholesterolinmg/100m1
1,259
1,540
4,512
3,383
1,145
2,925
4,271
2,977
1,424
1,511
4,182
3,404
1,090
3,040
3,963
2,903
170.6
171.3
171.9
174.0
166.7
173.0
172.4
173.8
172.7
177.5
179.2
179.5
174.4
177.3
179.1
179.0
1.72
2.34
0.86
1.34
2.13
1.27
1.10
1.11
2.50
2.44
1.09
0.97
5.16
1,21
1.18
1.56
771
1,122
4,055
3,280
818
2,237
3,913
2,822
928
1,102
3,760
3,258
784
2,337
3,600
2,752
165.6
167.4
171.5
173.5
164.1
170.4
171.5
173.5
171.8
175.1
178.7
179.2
173.4
176.0
178.7
178.7
2.83
2.60
0.98
1.38
2.49
1.52
1.11
1.04
3.46
2.74
1.17
1.03
6.90
1.57
1.20
1.65
487
402
424
87
322
671
335
132
495
400
407
117
302
679
363
127
178.5
180.9
175.2
189.4
172.9
181.6
181.5
178.5
174.4
183.7
183.5
185.8
177.2
181.9
182.4
180.7
3.06
4.58
2.89
5.66
2.70
2.49
3.22
4.97
2.85
3.42
3,22
5.32
4.70
2.98
3.54
4.96
NOTE: N = estimated number of youths in thousands; ~ = mean;s~ = standard error of the mean,
22
Table 13. Weighted sample size, mean cholesterol, and standard error of the moan for youths aged 12-17 years, by we, race, sex, and mmud family income: United States, 1966-70
Race, sex, and annual
famOy income
Cholesterol in mg/10D mlWhltn male
All bwomes . . . . . . .
Lussthan $3,000 . . . . . . .
$3,000.$4,999 . . . . . . . . . .
$5,000.$9,999 . . . . . . . . .
$l0,0000rmOre . . . . . . . .
Don’t know . . . . . . . . . . . . .
Blank or refused . . . . . . . .
White female
AH fncomes . . . . . . .
Lessthan $3,000 . . . . . . . . .
$3,W0.$4,999 . . . . . . . . . .
$6,000.$9,999 . . . . . . . . . . .
$lO,OQO OrmOre . . . . . . . . .
Don’t know . . . . . . . . . . . .
Blank orrefused . . . . . . . . . .
168.8.
155.2
165.6
169,4
172.2
162.2
173.7
176.0-
163.6
167.2
179.9
176.3
173.B
156.3
176.9.
174.5
192.3
171.9
171.2
171.2.
177.0.
166.7
174.9
192.0
172.7
166.9.
1,646
153
151
553
578
E4
46
1,594
169; 1.401,747 1.59 1.72S 1.594 166.:-
162;
162.,
166.(
169.[
167.,
173.4
1773
1.75[ 1.528180.2_
172,2
174.8
182.7
179.6
186,6
163.7
176.0.
173.2
176.8
177.4
179.9
180,3
187.0
167.2.
185,2
188.4
162.0206.4
!88.1.
I 75.4—
172.4
176.9
178.3
171.4.
,
173,5.
170.4
172.3
173.1
175.3
175.4
168.6
176.6.
160.9
176.4
180.2
176.5
179.2
172.0
183,1.
185.4
182.3
179.1.
.
.
1s5.5.
175.8
190.4
182.6
162.9.
.
1.26
9.80
6.84
2.28
1.90
4.62
4.66
1.82.
5.20
2.82
2.28
2.72
10.76
11.62
3.8o
6.86
4.13
7.91.
.
.
3.03
4.91
7.86
6.31
14.69.
.
1,686.
110
190
662
679
87
57
1,633.
138
211
661
535
51
36
256=
89
50
78
15
13
11
266—
75
69
57
19
21
6—
1.46.
2.73
4.73
2.24
2.63
7.32
8.14
1.67.
6.39
4.92
1.99
2.56
9.30
13.84
4.74.
5.16
12.95
8.80
13.71
6.61.
2.53.
3.80
7.09
9.29
20.08
9.82.
167.3.
157.8
167.0
165.9
172,1
171.3
153.5
176.6
1.&-
6.7:
3.71
1.9(
2.1/
3.54
6.5f
1.5:=
6.3f
3,51
2.25
2.72
5.7s
13.18
3.04.
6.20
5.17
7.73
14.29.
4.25
9.21
5.59
9.14
t 0.62.
.—
. . _111
209
786
567
41
43
1,665
7.71
3,86
1.94
2.04
6.66
11.23
1.23
127
1SC
681
606
58
76
1,667
129
211
632
501
67
55
1.542
3.69 142
5.23 181
2.11 540
2.62 459
6.84 37
6.92 66
2.02 1.502
175.5
162.[
169.<
170.s
162.E
174.t
184.2=
173.5
186.8
183.2
182.1
191.3
199.5
176.8
7.84
5.35
2.03
288
11.C4
8.26
1.23
9.62
8.24
2.72
2.26
6.14
46.12
6.44
180
195
641
571
63
24
280.
69
96
81
16
14
2
272.
77
76
81
19
16
4
5.46
5.42
2.66
2.17
5.28
43.42
3.01.
7.82
8.55
7.30
6.4o
42.62.
3.87
4.06
6.49
8.51
8.93●
.
160229
729
4B2
36
31
262.
85
60
78
9
6
3
275.
108
72
74
9
84
18:
207
611
506
69
27
241
66
65
59
17
12
235—
76
53
81
17
3
6
178.5
170.2
175.1
180.4
171.0
180.8
176.6.
176.7
172.3
170.2
189.6.
179.7
171.2
185.6
179.0
200.2●
.
156
113
589
571
69
44
231
179.f
165.2
176.5
179.:
177.5
193.C
173.2=.
166.9
168.E
178.0.
.
1s0.9.
177.6
189.3
180.2
183.8
179.2.
5.2n 111
4.02 146
3.16 529
3.77 591
4.26 86
12.13 40
4.46 225
Negro male
All incomes . . . . . . . . . .
Less than $3,000 . . . . . . . . .
$3,000.$4,899 . . . . . . . . . . .
$B,000.$9,899 . . . . . . . . . .$l0,0000rmOre . . . . . . . . .Don’t know . . . . . . . . . . . . .
Slankorrefumd . . . . . . . . . .
Negro female
Alllncamec . . . . . . . . . . .
Lets than $3,OOO . . . . . . . . .
$3,000.$4,999 . . . . . . . . . . .
$5,0Q0.$8,999 . . . . . . . . . . .
$l0,0000rmOre . . . . . . . .Don’t know . . . . . . . . . . . . .
Slankorrefwad . . . . . . . . . .
.72
54
8010
15
243.
89
52
57
23
14
7
10.81 85
3.84 58
7.40 48. 17. 12
6
2.46 237
182.0
177.7
154.3
187.2
163.6.
1s5.9
10.55
12.20
10.67
12.65
40.21.
4.27_2.57 71
8.01 60
5.48 58
13.58 30
8.08 6. 12
181.7
190.4
180.2
197.5.
190.9
5.66
9.75
7.42
11.59.
18.72—NDTE: N~estlmated numbrof youths inthousa"ds; ~=mean; s~=Sandard error of the mean.
23
Table 14. Wolghtedmmple size, mmncholesterol, andstandard error of themean forvouthsaged 12-17years, bvaw, mw, mx, and&u-tion ofwrenC United States, l966-7O
12 years I 13 years I 14 years I 15 yews I 16 year% I 17 yearsSex and education of parent
Meancholesterol inmg1100m1 .White male
Alledu@ ion groups .
Lessthan8 years . . . . . . . . .S-n years . . . . . . . . . . . . . . .12y.sars . . . . . . . . . . . . . . . .13years0rm0re . . . . . . . . .Unknown . . . . . . . . . . . . . . .
White female
Alleducatlongrourx .
Lessthan 8 years . . . . . . . . .S.ll years . . . . . . . . . . . . . . .12years . . . . . . . . . . . . . . . .13 yearsor more . . . . . . . . .Unkncr.+n . . . . . . . . . . . . . . .
Negro male
Alleduwdion groups .
Lessthan 8 years . . . . . . . . .S-n years . . . . . . . . . . . . . . .12years . . . . . . . . . . . . . . . .13 years or more . . . . . . . . .Unknown . . . . . . . . . . . . . . .
Negro female
All education groups .
Lessthan 8 years . . . . . . . . .%lly ears. . . . . . . . . . . . . . .12y0ars . . . . . . . . . . . . . . . .13 yearsor more . . . . . . . . .Unknown . . . . . . . . . . . . . . .
180.2 1.59 173.5 1.26 1,586 188.8
145 169.8341 167.8632 VX3.6523 172.2
14 .
1.633 175.0
107 166.3429 175.9548 174.6440 176.0
9 ●
256 176,9
27 16S.6130 183.6
Ed 173.920 170.311 “
25$ 177.0
167.3! 1.46 1,594 166.7 1,628 169.7 1.40-
8.342.872.632.07
11,41
1,23
I ,747.
119412675528
13
1,886.
lt%379650454
28
230-
57130
5332
3
272
1.728 1.46.
6.443.081.462.21
.
1.67
6.932.542.363.87
.
4.74.
39.5a7.949.17
12.63.
2.53.
3.453.856.37
13.40.
1,546.
139280652461
14
I .594.
127479578391
20
241.
5710358
63
235.
47109
63143
—
1.76.
4.962.403.132.185.44
2.02.
9.932.922.754.50
28.73
4.46.
6.6110.28
7.4311.79
.
2.46
. . .173.8176.7160.3162.8
.
176.0.
173.3174.9179.5180,3177.3
187.2.
184.6193.51S0.8175.4
.
175.4
5.673.062.312.95
.
1.23-
7.043.062.152.14
16.77
3.01.
5.257.435.11
10.08●
3.87
104381713501
19
1,557.
182266661452
16
262.
711026423
3
276.
46147
8511
5
162.5176.2173.5174.2177.4
176.6
172.2174.6160.2174.4
.
163.1=
173.1183.4192.7182.3
.
186.5
3.043.722.432.479.75
1.62.
8.012.262.462.56
.
3.80.
7.W4.928.798.06
.
3.03
+
160.5 5.85164.3 3.49lm.5 1.92158.2 2.24149.5 13.96
176.5 1.58
170.1 9.25176.8 2.58179.0 2.62174.6 1.43
. .
175.6 3,04
163371E80446
28
1,542
106349572490
24
231
161.3167.3167.4167.9166.6
177.7.
176.1171.3176.0183.3200.7
173.3
1433426322a3
49
1,602-
96344491615
56
225
188.3167.6l&3.8173.0175.0
184.2
166.4182.3184.1181.9217.2
176,8
19.162.422.462.48
23.21
6.44. .46
1034627
7
243
173,8155.9182.3163.7
.
180.9
621043123
6
237
170,6183.1169.6178.1
.
165.9
7.777.71
10.4017.64
.
4,27
165.0 4.43176,8 4.06184.0 6.53
. .
. .
179.7 4.25
4696793515
177.2176.8173.3lffl.5175.0
2.067.219.407.32
12.16
175.6165.1192.2
,<
13.865.575.13
.
. -L53 173.1
186 1s0.571 176.525 171.512 “ ---L
178,0 4.41178.1 7.27180.4 6.45196.5 16.20
. .
61104501810
173.5179.3187.7
.
.
6.593,826.79
.
.—
50117352412
—
167.0187.5192.7173.4
.
12.M6.76
11.767.27
●
NOTE: N=estlmated numberof youths lnthoumnd%; ~=mean; s~=flandard error of the mean.
24
Table 15. Mean cholesterol levels and standard deviations of the population1960-62
distribution foraduks, by sex and age: Unitad States,
Age
18-24 years25-34 yaars35-44 yaars45-54 years55.64 yaars65-74 years75-79 years
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
., ...,.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..,. .,..,. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
lMgper~OOml.NOTE; X=mean; sx=standard deviation.
Men
178.1205.9225.8230.5232.8229.5224.5
Sx
40.744.649.445.649.047.348.7
Women
184.7197.9213.6236.8262.3265.7245.3
Sx
47.941.945.350.063.058.865.7
APPENDIX I
TECHNICAL NOTES ON METHODS
The Survey Design
The sampling plan of Cycle III of the HealthExamination Survey followed a multistage, strat-ified probability sample of clusters of house-holds in land-based segments in which a sampleof the U.S. population (including Alaska andHawaii) aged 12 through 17 years was selected.Excluded were those youths confined to institu-tions and those residing on any of the reserva-tion lands set aside for use by American Indians.
The sample design of Cycle III is similar tothat of Cycle II in that it uses the same 40sample areas and the same segments. The deci-sion to incorporate this feature into Cycle IIIwas not made prior to the selection of the CycleII sample, although it is consistent with theinitial concept of a single program for persons6-17 years old. The final decision to use thisidentical sampling frame was made during theoperation of the Cycle II program.
The successive elements for this sample designare primary sampling unit; census enumerationdistrict; segment (a cluster of households);household; all eligible youths; and finally, thesample youth. Every eligible youth within thedefined population has a kfiown and approxi-mately equal chance for selection into thesample.
The steps of drawing the sample were carriedout jointly with the U.S. Bureau of the Census;the starting points were the 1960 decennialcensus lists of addresses and the nearly 1,900primary sampling units (PSU’S) into which theentire United States was divided. Each PSU is astandard metropolitan statistical area (SMSA), acount y, or a group of two or three contiguouscounties. These PSU’S were grouped into 40strata so that each stratum had an average size of
about 4.5 million persons. This grouping maxi-mized the degree of homogeneity of the PSU’Swithin each strata with regard to the populationsize, degree of urbanization, geographic proxim-ity to one another, and degree of industrializa-tion. The 40 strata were then classified into fourbroad geographic regions of 10 strata each and,within each region, cross-classified by four popu-lation density classes and by the rates ofpopulation change from 1950 to 1960. Using amodified Goodman-Kish controlled-selectiontechnique, one PSU was drawn from each of the40 strata.
The sampling within the PSU’S was carriedout in several steps. The first step was theselection of census enumeration districts(ED’s)–small, well-defined areas of about 250housing units. The entire nation was divided intoED’s for the 1960 population census, and eachED was assigned a “measure of size” equal tothe rounded whole number resulting from a“division by nine” of the number of childrenaged 5-9 in the ED at the time of the 1960census. A sample of 20 ED’s in the sample PSUwas selected according to a systematic samplingtechnique, with each ED having a probabilityy of,selection proportional to the population ofchildren aged 5-9 years at the time of the 1960census date. From each ED a random selectionof one measure of size (segment) was taken.
Minor changes required in the Cycle III designwere (1) that it be supplemented for newconstruction to a greater extent than had beennecessary in Cycle II and (2) that reservesegments be added. Although it was the plan forCycle III to use the Cycle II segments, it wasrecognized that within several PSU’S additionalreserve segments would be needed to avoid the
26
risk of having an insufficient number of exam-inecs. This was prompted by the fact that fourof the PSU’S in Cycle II had yields of less than165 eligible children and several others weremarginal in their yield. In addition, there was a3-year interval between Cycle II and Cycle III,so that it was quite possible” for some segmentsto have been completely demolished due tohighway construction or urban redevelopment.
The time available for examinations at aparticular location, or stand, is necessarily set farin advance of any preliminary field work at thestand. Therefore, the number of examinationsthat can be performed at a particular location isdependent on the number of examining daysavailable. At the majority of locations, thenumber of days available, excluding Saturdays,is 17. At the rate of 12 examinations each day,this provides for 204 examination slots. Exami-nations are conducted on Saturdays if necessary.Because of rescheduling for cancellations orno-shows, the maximum number of youths thatis considered for inclusion in the sample is 200.When the number of eligible youths exceeds themaximum, subsampling is performed to reducethe number to manageable limits. This is accom-plished through the use of a master list, which isa listing of all eligible youths in order bysegment, serial number (household order withinsegment), and column number (order in thehousehold by age). After the subsampling ratehas been determined, every nth name on the listis deleted, starting with the yth name, y being arandomly selected number between 1 and n.Youths who are deleted from the Cycle 111sample but who were examined in Cycle II andany twin who may have been deleted arescheduled, if time permits, for an examinationto be included only in the longitudinal studyportion or twin study portion of the survey.Their data are not included in the report as partof the regular sample.
Since the strata are roughly equal in popula-tion size and a nearly equal number of sampleyouths were examined in each of the sample “PSU’S, the sample design is essentially self-weighting with respect to the target population;that is, each youth 12 through 17 years old hadabout the same probabilityy of being drawn intothe sample.
The adjustment upward for nonresponse isintended to minimize the impact of nonresponseon final estimates by imputing to nonrespond-ents the characteristics of “simiku-” respondents.Here “similar” respondents were judged to beexamined youths in a sample PSU having thesame age (in years) and sex as those notexamined in that sample PSU.
The poststratified ratio adjustment used inCycle III achieved most of the gains in precisionthat would have been attained ‘if the sample hadbeen drawn from a population stratified by age,color, and sex, and it made the final sampleestimates of population agree exactly with in-dependent controls prepared by the U.S. Bureauof the Census for the noninstitutional popula-tion of the United States as of March 9, 1968(approximate midsurvey point) by color and sexfor each single year of age 12 through 17. Thesampling weight of every youth examined ineach of the 24 age, race, and sex classes isadjusted upward or downward so that theweighted total within the class equals the inde-pendent population control.
A more detailed description of the samplingplan and estimation procedures is included inVital and Health Statistics, Series 2-Number43,38 and in Series l-Numbers 11, 53, and 8A,which describe the plan and operation of thefirst three cycles of the HeaIth ExaminationSurvey.
Some Notes on Response Rates
As mentioned previously, the sample designsof the second and third cycles of the HES weresimilar. Differences did occur, however, in re-sponse rates of various subgroups of thesesamples, and these differences deserve someconsideration here.
Most importantly, the number of youthsselected for examination increased from 7,417in Cycle II to 7,514 in Cycle HI. The responserate—i.e., the number of youths selected whowere actually examined—decreased from 96 per-cent in Cycle II to 90 percent in Cycle III. Ofthe youths examined in Cycle II, 13.9 percentwere Negro, compared with 14.8 percent of
Note.–A list of references follows the text.
27
those examined in Cycle III. This differencedoes not reflect a difference in the percentage ofNegro youths selected for examination, butrather, a smaller decrease in response rate forNegro youths between the two cycles than wasthe case for white youths. In actuality, 13.8percent of the sample selected for examinationwas Negro in Cycle III, corresponding to 13.5percent in Cycle II. However, whereas theresponse rate for white youths dropped from95.6 percent in Cycle II to 89.1 percent in CycleIII, the response rate for Negro youths droppedfar less, from 98.4 percent to 96.6 percent.Thus, relatively better response from the Negroportion of the sample in Cycle III increased theirpercentage of actual examinations as comparedwith the previous cycle.
Examination of sample sizes in this reportclearly shows that at every age group, fewerfemales than males were actuaIly examined.This, again, is not attributed to differences innumbers of youths selected in the samplingdesign, but rather to the following differentialresponse rates between males and females:
Age Male Female
Total . . . . . . . . . . . . . . . . . . . . . . 91.4 88.7
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
Note that at each age group the response rate formales exceeded that for females.
A similar analysis of response rates can bedone by age, race, and sex, as follows:
AgeWhite Negro White Negro
male male female female
Total . . . . . . . . . 90.5 97.6 87.4 95.8
12 . . . . . . . . . . . . . . . 92.6 99.0 90.1 98.913 . . . . . . . . . . . . . . . 92.5 98.8 91.1 96.814 . . . . . . . . . . . . . . . 91.0 97.8 89.6 96.215 . . . . . . . . . . . . . . . 90.7 97.7 86.4 98.616 . . . . . . . . . . . . . . . 89.2 95.0 86.6 93.017 . . . . . . . . . . . . . . . 86.5 95.8 80.2 91.4
The above clearly indicates that for all agesunder consideration in Cycle III of the HES, theresponse rate for Negro youths exceeded thatfor white youths of the same sex and age.
Reasons for differences in response rates aremany, but may range from the incentive to getexamined in order to miss a day of school, tofear of the examination itself, to inhibitionswith respect to being examined. The worstresponse rate was recorded for the oldest fe-males, that is, those aged 17 years.
Parameter and Variance Estimates
Because each of the 6,768 sample children hasan assigned statistical weight, all estimates ofpopulation parameters presented in HES publi-cations are computed taking this weight intoconsideration. Thus, ~, the estimate of a popula-tion mean p is computed as follows:
n
xz wjx~j=1
= ZWj
where Xi is the observation ortaken on the ith person and Wi isweight assigned to that person.
measurementthe statistical
The HES has an extremely complex samplingplan, and obviously, by the very nature of thesample, the estimation procedure is complex as
well. For estimating the reliability of findings, amethod is required that “reflects both the lossesfrom clustering sample cases at two stages andthe gains from stratification, ratio estimation,and poststratification.”8 9
The method for estimating variances in theHES is the half-sample replication technique.The method was developed at the U.S. Bureauof the Census prior to 1957 and has at timesbeen given limited use in the estimation of thereliability of results from the Current PopulationSurvey. This half-sample replication technique isparticularly well suited to the HES because thesample, although complex in design, is relatively
small (6,768 cases) and is based on but 40 strata.This feature permitted the development of avariance estimation computer program that pro-duces tables containing desired estimates of
28
aggregates, means, or distributions, togetherwith a table identical in format but containingthe estimated variance of the estimated statis-tics. The computations required by the methodare simple, and the internal storage requirementsare well within the limitation of the IBM 360-50computer system used at the National Center forHealth Statistics.
Variance estimates computed for this reportwere based on 20 balanced half-sample replica-tions. A half sample was formed by choosingone sample PSU from each of 20 pairs of samplePSU’S. The composition of the 20 half sampleswas determined by an orthogonal plan. Tocompute the variance of any statistic, thatstatistic is computed for each of the 20 halfsamples. Using the mean, xi, as an example, theweighted mean of the entire undivided sample(~) is computed. The variance of the mean is themean square deviation of each of the 20half-sample means about the overall mean.Symbolically,
20i~l (~. - 5?)2
Var(x)= Z.
and the standard error of the mean is the squareroot of Var(~). In a similar manner thestandard error of any statistic may be computed.
A detailed description of this replicationprocess has been published in Vital and HealthStatistics, Series 2-Number 14.39
Standards of Reliability and Precision
All means, variances, and percentages appear-ing in this report met defined standards beforethey were considered acceptably precise andreliable.
The rule for reporting means and percentilesconsisted of two basic consecutive criteria: that
a sample size be at least five; and that theestimated coefficient of variation (i.e., the stan-dard -grror of the mean divided by the mean, or(SX /X) be less than 25 percent. Thus, if thesample size was too small, or if, given adequatesample size, the variation with respect to themean was too large, then the estimate wasconsidered neither precise nor reliable enough tomeet the standards established for publication.
Imputation
In addition to the subject nonresponse dis-cussed above, the problem of item nonresponsemerits consideration here. In this situation,information about a respondent is completewith the exception of a missing cholesterolvalue.
A regression method was initially chosen forthe imputation of missing cholesterol values. Ifother variables could have been found whichcorrelated with cholesterol, this other informa-tion could have been effectively used in aregression scheme of imputation. Two “bestpossible” predictor variabIes were chosen intriceps and subscapular skinfolds. After stand-ardizing these two variables by age, race, and sexgroup, a regression was run. However, theresultant multiple R of O.1316 was not con-sidered to be sufficiently high to justify aregression approach to replacing missing choles-terol values.
Instead, the technique of replacing a missingdatum by the value recorded for a randomlyselected respondent of the same age, sex, andrace was employed. When only one of the twocholesterol values was missing, the recordedmeasurement was substituted for the unknownvalue. This occurred in only one instance.Imputation where there was no value recordedfor examinees was necessary in 175 cases (i.e.,for 2.6 percent of total respondents).
Uuo
29
APPENDIX II
DEMOGRAPHIC VARIABLES
Regional and demographic characteristics bywhich the population has been classified for thisreport are defined as follows:
Age and sex. –Population was classified into12 age-sex groups-the six ages 12-17 years bysex. Birth certificates verified the age of 92percent of the youths. Age stated by the parentswas accepted as the true age for the other 8percent. Age is expressed as years attained at lastbirthday.
Race. –Serum cholesterol value was reportedby race for white and Negro youths. Youths ofother races were not sampled sufficiently forcomparison purposes and represented only 0.55percent of the sample.
Re.~-on.–Regional data are presented for fourA
regions of the continental United States.
Re~.on
Northeast
Midwest
South .
. . .
. . .
. . .
States Included
Maine, Vermont, Newshire, Massachusetts,Island. Connecticut.
Hamp-Rhode
NewYork, Pennsylvania, New Jer-seyMinnesota, Wisconsin, Michi-gan, Iowa, Missouri, Illinois,Indiana, OhioDelaware, Maryland, Virginia,District of Columbia, WestVirginia, Kentucky, Tennes-see, North Carolina, SouthCarolina, Georgia, Florida,
West . . . . .
Alabama, Mississippi, Arkan-sas, LouisianaWashington, Oregon, Idaho,Montana, North Dakota,South Dakota, Wyoming,Nebraska, Kansas, Colorado,Utah, Nevada, California, Ari-zona, New Mexico, Texas,Oklahoma, Alaska, Hawtil
Family income. –The income recorded wasthe total income received during the past 12months by the head of the household and allother household members related to the head byblood, marriage, or adoption. This income wasthe gross cash income (excluding pay in kind)except in the case of a family with its own farmor business, in which case net income wasrecorded.
Education of parent or guardian.–This itemwas recorded as the highest grade that had beencompleted in school. The only grades countedwere those that had been completed in a regular
school in which persons were given formaleducation: graded or private schools, either dayor night schools, with either full-time or part-time attendance. A “regular” school is one thatadvances a person toward an elementary or highschool diploma, or a college, university, orprofessional school degree. Education in voca-tional, trade, or business schools outside theregular school system was not counted in deter-mining the highest grade of school completed.
000
APPENDIX Ill
TECHNIQUES OF MEASUREMENTAND QUALITY CONTROL
MEASUREMENT
The HES Blood-Drawing Technique and ItsHistorical Development
For a variety of reasons, it was decided not to‘ attempt to draw blood specimens from children
6-11 years of age in Cycle II. The childrenincluded in the national probability sample camefrom all regions of the country and all culturaland socioeconomic groupings and ways of life. Itwas assumed, therefore, that some had neverbeen to a physician before and that otherswould have very bad memories of and associa-tions with such visits. In addition, because thesample covered the entire spectrum of behav-ioral and physical development extant in theUnited States, severe technical and behavioralproblems resulting from immaturity of 6- and7-year-olds would be likely. It was believed thatfear or extreme distaste for having a bloodsample drawn on the part of many potentialsubjects might severely affect the response rate,which is so crucial to a survey like this. When itis remembered that the overall response rate forCycle II was a remarkable 96 percent, it isdifficult to argue with this line of reasoning.Now, of course, it can never be known howgreat a diminution of the response rate wouldhave been caused by the inclusion of a bloodsample from younger children in Cycle II.
In the early planning stages of Cycle III, itwas decided to obtain a blood specimen fromthe youths 12-17 years if at all feasible, that is,primarily if the price in terms of diminishedresponse rate and cooperation of the examinees
would not be too high. Accordingly, in threeseparate pretests, investigations were conductedregarding the problem of developing a satisfac-tory blood-collection technique for this agegroup and examination setting. What was desiredwas to draw the optimum amount of blood,without causing emotional upset to the exam-inee, or without affecting his performance inany of the procedures to follow. The amount ofusable blood that could “be drawn posed alimiting factor on the number of blood chemis-try tests that could be performed and greatlyinfluenced the acceptance or rejection of anentire possible area of’ the examination, such asthe nutritional assessment. Logistical problemsalso had to be resolved involving the handling,separating, and packaging of drawn blood sothat there would be a minimum of blood lossand packaging error. For the refrigerated butunfrozen blood, time from shipment to deliverywas critical; therefore, arrangements had to bemade with postal authorities to assure promptdelivery to the laboratories in order to avoidspoilage.
There was a trial-and-error process, and therewas good advice and help from many sources indeveloping a satisfactory blood-drawing tech-nique. The chief sources of help, outside of theimmediate HES technical staff, were Dr. WilmaBias and Dr. Bernice Cohen of The JohnsHopkins University; Dr. Gerald Cooper, Chief ofLaboratories, Center for Disease Control, PublicHealth Service, Atlanta, Ga.; the many teenagesubjects during our pretest who gave valuablesuggestions and who pointed out, either as overtadvice or by their immediate reactions, specific
31
points to be avoided; and, finally, the profes-sional and technical division of the Becton-Dickinson Company, Rutherford, N.J. Thelatter, through several personal visits by arepresentative of their technical division, notonly gave excellent technical advice on blood-collection techniques and the use of alternativeequipment, but also devised a special fitting thatmade the transfer from one vacutainer tube toanother much smoother.
During the pretesting, it was learned thatmany subjects did not like to see any part of theblood-drawing procedure, including their ownblood in tubes. Therefore, a technique wasemployed that minimized the subject’s attentionto the operation. Effective screening wasachieved by having the subject lie down and bydraping and keeping the arm and tubes wellbelow the level of the examination table. Afterthe skin area was cleansed with alcohol, theblood was drawn from the antecubital fossa bythe physician-nurse team. At the discretion ofthe physician, a tourniquet was used to fill thevein; however, once the needle was inserted intothe vein, the tourniquet was taken off the armso that the blood flowed freely.
A B-D blood culture needle and tube wereused to draw blood. Using the specially preparedlink fitting, the nurse inserted the short needleinto a vacutainer tube holder. The tube wasclamped with a hemostat until the vein waspunctured and the vacutainer was inserted intothe holder.
From the one free-flowing venipuncture, atotal of only 55 cm3 of blood was collected infour separate vacutainer tubes from all male andalmost half of the female subjects. (The differ-ence was that all males had a separate specimendrawn to be frozen and stored as plasma forfuture testosterone determination and almosthalf of the females provided a replicate bloodspecimen for quality control of the laboratorydeterminations; the remaining females had 40cm3 drawn.)
Each test tube was labeled with the exam-inee’s number and left in the test-tube holdingrack at room temperature for 1 hour. The nursethen placed the tubes in the laboratory refrigera-tor, along with 10 extra examinee identificationlabels for use by the technicians.
An analysis that attempts to estimate theimpact of the addition of a blood sample on theCycle III sample response rate is in progress.
Cholesterol Chemistry Determinations
The serum samples were shipped to the LipidStandardization Laboratory at the Center forDisease Control in Atlanta, Ga., where they werekept frozen until assay.
After being thawed, samples were divided andstored in two vials. All sample vials were thenrandomized over a 6-day period so that a pairof duplicate samples might have been analyzedon the same day or as many as 6 days apart.
A semiautomated method based on theAbelI-Kendall procedure was employed to meas-ure total cholesterol. Determinations werecarried out on an assembly-line basis so that twoanalysts could analyze more than 120 samples induplicate per day.
MONITORING SYSTEMS
In addition to the sampling considerationsalready discussed, the quality of data collected isalso a special concern. One of the main purposesof the monitoring system employed in thesurvey was to indicate whether the measure-ments produced by our measurement processattained the desired quality. A second majorpurpose was to make possible quantitative sum-mary descriptions of residual measurementerrors to aid in the interpretation of survey data.
The monitoring system as applied to thetaking of blood samples consisted of a formalsystem of replicate examinations (described laterin this appendix). Replicate measurements areuseful for a variety of reasons; for example, as ameans of increasing precision of estimates ofindividual measurements, as a training tech-nique, and as a monitoring system that includesthe objective of overall evaluation of measure-ment errors. These objectives are not incompat-ible, and replicate data collected primarily forone of these objectives often indirectly, if notdirectly, accomplish one or both of the remain-ing two. For this reason replicate data are mostoften collected with a combination of theseobjectives in mind.
32
Methods of Taking Replicate Measurements
A major source of uncertainty in estimatesderived from replicate measurements is in theinability to make the replicate measurementunder precisely the same conditions and in thesame manner as the original measurement. Thisuncertainty is difficult to evaluate, and mostattempts are restricted to subjective statementsconcerning the direction and/or size of the biasand the need for concern in the analysis of data.
In this study two extra blood samples weredrawn from a subsample of Cycle III examineesand were included in the shipment sent forprocessing to the Lipid Standardization Labora-tory at the Center for Disease Control inAtlanta, Ga. Replicate blood samples were takenfrom the same venipuncture as the regular bloodsamples and, from shipment to final processingat the CDC, were treated the same as the originalsamples. These replicates were labeled withdummy examination identification numbers andwere recorded by the nurse in a replicate logbook. All samples were submitted for laboratoryanalysis with no indication that any two samplescame from the same examinee. Each replicatewas analyzed as usual, where each sample wassplit by the technician and duplicate determina-tions were performed; if a difference greaterthan 9 mg/100 ml was found (it was usually areading or recording error), a second set ofduplicate determinations was performed. Theanalyses of replicates and originals were per-formed under identical laboratory conditions bythe same technician in a true double-blindmanner.
Cycle III examinees were chosen systemati-cally for replicate blood determinations. Onevery third day for the first 15 days of eachexamination center, two extra blood sampleswere drawn (preferably from girls) for thereplicate study. Since in a voluntary survey it isimpossible to follow a statistically random proc-ess in scheduling subjects, the replicate desi~did not ensure that its subjects would be“representative” of those in the larger Cycle HIHealth Examination Survey. It is felt, however,that this is not a crucial issue since the matter ofconcern in undertaking the replicate study is notthe determination of possible differences in the
values of the measurements, but rather thedetermination of possible differences in errorsassociated with the measurements.
Results of the Replicate Study for Cholesterol
Two readings were made of the original splitsample obtained for each of the 6,592 exam-inees (97 percent of the total sample) by thesame technician. An extra blood sample wasdrawn for replicate studies on 424 examinees, ofwhich 98 percent were adequate.
Frequency and percent distributions of theabsolute differences between the duplicateddeterminations on the original specimen and alsobetween the duplicated determinations on thereplicate and original specimens are presented intable I. The first two columns represent thedifferences between the duplicated determina-tions on the original specimen; the second twocolumns represent the differences between thefirst recorded determination values of the origi-nal specimen and that of the replicate specimen;and the third two columns represent differencesbetween the second (or duplicated) recordedwdues of the original and replicate specimens.
As a summary statistic of the distribution ofdifferences between replicate and originalcholesterol determinations presented in table I,we have computed V, the percentage technicalerror of measurement which is given by
where
n
d:
x
is the number of pairs of measurements inthe study,
is the square of the difference betweenmembers of the ith pair of measurements(i=l,...,n), and
is the arithmetic mean of the 2n measure-ments in the study.
The percentage technical error, V, can be inter-preted as a “coefficient of variation” and is a
33
Table 1. Fraquancy and parcent distribution of absolute differences betwean duplicated determinations of original specimans and
batwaan original and replicated specimans
Absoluta difference
(mg/100 ml)
Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .lo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43 . . . . ..-. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
58 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
79 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
82 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Difference betwean
duplicated determina-
tions of original
specimans
Frequency
6,592
996
1,285
1,037881752
510
390
308
225
14143
1
4
4
1
1
2
1
h-
1
2
1
1
1
1
1
1
1
of total
100.0
15.1
19.5
15.713.417.4
7.7
5.9
4.7
3.4
2.10.7
0.10.1
.-
Difference betwaan original and replicated
specimens
First determinations
Frequency
408
32
46
57405234
14
21
2816
3
5
14
6
4
4
2
523
1
3
11
1
1
1
1
11
1
11
1
1
1
1
Percent
of total
100.0
7.8
11.3
14.09.8
12.78.3
3.4
5.1
7.1
3.90.7
1.2
3.4
1.5
1.0
1.0
0.5
1.20.5
0.7
0.2
0.7
0.20.2
0.2
0.2
0.2
0.2
0.20.20.2
0.20.2
0.2
0.2
0.2
0.2
jecond determinetions
Frequancy
407
35
46
374342
34
25
27
22
1715
12
11
6
5
6
6
2
1
1
1
1
1
2
1
21
1
1
1
1
1
Percent
of total
100.0
8.6
11.3
9.110.610.3
8.4
6.1
6.6
5.44.23.7
2.9
2.7
1.51.2
1.5
1.5
0.5
0.2
0.2
0.2
0.2
0.2
0.5
0.2
0.50.2
0.2
0.2
0.2
0.2
0.2
34
dimensionless constant. It essentially describesthe size of measurement error relative to themean value of a measurement. As one measureof the differences between specified determina-tions, the values of V are given below.
Specified Determinations v
First determinations on origi-nal and replicate specimens 4.7 percent
Second determinations onoriginal and replicatespecimens 4.6 percent
Duplicated determinations onoriginal specimen 1.9 percent
The reason for the contrast in V in the figuresabove is that the original split sample wassubject to a repeat analysis if the readingsdiverged by more than 9 mg percent. Cholesterolvalues were not as easily replicated at the Centerfor Disease Control as uric acid values hadbeen.g When the cumulative frequency andpercent distributions in table I are examined, itis found that more than 15 percent of each ofthe two distributions on the right side of thetable are above the 9 mg/100 ml level. However,the fact that the coefficient of variation is less
than 5 percent in each case above does attest tothe overalI reproducibility of cholesterol.
Data Handling Verification
Quality control considerations were not con-fined to the laboratory. Data were subject to thepossibility of error every time a human handtouched a keyboard or moved a pencil across apage. After the data had been put on punch-cards, they were transferred to magnetic tape.Subsequent handling of the data by pro-grammers and transcription by clerks providedother sources of error.
To verify all operations, the entire cohort of16-year-old Negro maIes was subjected to athorough independent manual audit. The com-puter tape printout of each subject’s serumcholesterol value was listed opposite his identifi-cation. These were individually checked againstthe values originally recorded at the Center forDisease Control. The mean was then computedmanually on the desk calculator and provedidentical to the computer’s mean value.
In addition, all age, race, and sex identifica-tion for members of this cohort were found tocorrespond exactly to the information given onthe household interview forms completed bycensus interviewers.
000
*U.S. GOVERNMENT PRINTING OFFICE: 1976– 21O-9SLS7
35
Series 1.
Series2.
VITAL AND HEALTH STATISTICS PUBLICATIONS SERIES
Formerly hblic Health Service Publicatwn No. 1000
Programsand Collection l+ocedures.-Reports which describe the general programs of the NationalCenter for Health Statistics and its offices and divisions, &ta collection methods used, definitions, andother material necessary for understanding the data.
Data Evaluationand Methods Research.-Studies of new statistical methodology including experimental
Sm”es3.
Series4.
tests of new survey methods, studies of vital statistics collection methods, new analytical techniques,-. -.
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 a,~d Comms”ttee Reports. –Fd reports of major committees concerned with vital andhealth statistics, and documents such as recommended model vital registration laws and revised birthand death certitkates.
Series 10. Dataporn the Health Interview Survey.-Statistics on illness; accidental injuries; disability; use ofhospital, medical, dental, and other services; and other health-related topics, based on data collected ina continuing national household irkterviewsurvey.
Sert”es11. Dataporn the Health Examination Survey.-Data from direct examination, testing, and measurementof national “samplesof the civilian, noninstitutionalized population provide the basis for two types ofreports: (1) estimates of the medically detlmed prevalence of ipecitlc diseases in the United States andthe distributions of the population with respect to physical, physiological, and psychological charac-teristics; and (2) analysis of relationships among the various measurements without reference to anexplicit ftite universe of persons.
Series 12. Datafrom the InstitutionalizedPopulation Surveys.–Discontinued effective 1975. Future reports fromthese surveys will be in Series 13.
Sers”es13. Data on Health Resources Utikration.–Statistics on the utilization of health manpower and facilitiesproviding long-term care, ambulatory care, hospital care, and family planning services.
Series 14 Data on Health Resources: Manpower and Facilities.–Statistics on the numbers, geographic distrib-ution, and characteristics of health resources including physicians, dentists, nurses, other health occu-pations, hospitals, nursing homes, and outpatient facilities.
Series 20. Data on Mortality.-Various statistics on mortality other than ss included in regular snnual or monthlyreports. Special analyses by cause of death, age, and other demographic variables; geographic and timeseries analyses; and statistics on characteristics of deaths not available from the vital records, based onsample surveys of those records.
Sers”es21. Data on Natality, Marn”age,and Divorce.–Various statistics on natality, marriage, and divorce otherthan as included in regular annual or monthly reports. Special analyses by demographic variables;geographic and time series analyses; studies of fertility; and statistics on characteristics of births notavailable from the vital records, based on sample surveys of those records.
Series 22. Datufrom the NationalMortality and Natality Surveys.–Discontinued effective 1975. Future reportsfrom these sample surveys based on vital records will be included in Series 20 and 21, respectively.
Serz”es23. Data j+om the Nats”onalSurvey of Family Growth. –Statistics on fertility, family formation and disso-P lution, family planning, and related maternal and infant health topics derived from a biennial survey of<.+
., a nationwide probability sample of ever-married women 15-44years of age.
For a list of titles of reports published in these series, write to: Scientific and Technical Information BranchNational Center for Health StatisticsPublic Health Service, HRARockville, Md. 20852
POSTAGE AND FEES PAID
U.S. DEPARTMENT OF[’. S DE PARTMENT OF HEW
HEALTH, EDUCATION, AND WELFAREPublic Health Service HE W390Health Resources Administration
5600 Fishers Lane THIRD CLASS
Rockville, Md. 20852 BLK. RATE
OFFICIAL BUSINESSPenalty for Private Use, $.300
![Original Article Effects of G-CSF on serum cholesterol and ...dase or lowering serum cholesterol [16-20]. Administration of M-CSF can also lower serum cholesterol levels both in animal](https://static.fdocuments.in/doc/165x107/608dd04a2d10932c166a55b2/original-article-effects-of-g-csf-on-serum-cholesterol-and-dase-or-lowering.jpg)
