Citation
The interrelationships between birth weight, infant mortality and altitude in Colorado

Material Information

Title:
The interrelationships between birth weight, infant mortality and altitude in Colorado
Creator:
Unger, Cynthia Beth
Publication Date:
Language:
English
Physical Description:
ix, 69 leaves : illustrations, map ; 29 cm

Subjects

Subjects / Keywords:
Newborn infants -- Mortality -- Colorado ( lcsh )
Birth weight, Low -- Colorado ( lcsh )
Birth weight, Low ( fast )
Newborn infants -- Mortality ( fast )
Colorado ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 60-63).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Department of Anthropology.
Statement of Responsibility:
by Cynthia Beth Unger.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
16855248 ( OCLC )
ocm16855248
Classification:
LD1190.L43 1986m .U53 ( lcc )

Full Text
(
THE INTERRELATIONSHIPS BETWEEN BIRTH WEIGHT,
INFANT MORTALITY AND ALTITUDE IN COLORADO
by
Cynthia Beth Unger
B.S. Miami University, 1973
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Arts
Department of Anthropology
1986


This thesis for the Master of Arts degree by
Cynthia Beth Unger
has been approved for the
Department of
' Anthropology
by
Janet Moone
Date
j S36


Ill
Unger, Cynthia Beth (M.A. Anthropology)
The Interrelationships Between Birth Weight, Infant
Mortality and Altitude in Colorado
Thesis directed by Associate Professor Lorna G. Moore
Birth weight, an important predictor of infant
survival, is known to decrease with increasing alti-
tude. This study sought to determine the contribution
of reduced birth weight to infant mortality over the
range of altitudes present in Colorado. The computer-
ized birth and death certificates of 184,932 survivors
and 1,684 infant deaths in Colorado occurring between
1979 and 1982 were analyzed. Whereas previously com-
piled data showed increased infant mortality at high
altitudes (>9,000 ft.) in Colorado, the 1979-1982
rates indicate that infant mortality at high altitude
has decreased to levels comparable to those at -lower
altitudes in the state. The incidence of low birth
weight infants (<2500g) at high altitude, has declined
at a faster rate than at lower altitudes, although
birth weight continues to correlate negatively with
altitude. Increased mortality rates among smaller
babies occur at all altitudes, but a greater propor-
tion of 1500-2500g infants survive in high (>9,000 ft.)
as compared to low (<7,000 ft.) altitude regions (95.5
vs. 92.2%, P=.05). Birth weight specific mortality
increased among males at high altitude, but decreased


iv
among females. The recent improvements in infant sur-
vival at high altitude in Colorado may result from
advances in medical care, increased ability to trans-
port high risk infants to facilities specializing in
their care, an alteration in the association between
infant birth weight and mortality at high altitude, or
other unidentified factors.


V
CONTENTS
CHAPTER
I. INTRODUCTION.................................. 1
II. BACKGROUND.................................... 3
Human Adaptation............................ 3
Birth Weight as an Indicator of
Survivorship............................... 5
Birth Weight and Infant Mortality at
High Altitude............................. 6
III. RESEARCH OBJECTIVES......................... 9
IV. METHODOLOGY.................................. 10
Site Selection.............................. 10
Data........................................ 11
Definitions................................. 12
Statistics................................ 15
V. RESULTS.................................... 17
Historical Perspective...................... 17
Relationships Between Altitude, Birth
Weight and Gestation...................... 21
Relationships Between Altitude,
Mortality and Gestation................... 31
Relationships Between Altitude, Birth
Weight, Mortality and Gestation........... 31
Level of Care............................... 39
)
Complications............................ 41
Cause of Death.............................. 44


vi
CONTENTS (Continued)
CHAPTER
Mother's Altitude of Birth................. 44
Socioeconomic Factors...................... 44
Urban vs. Rural......................... 49
Maternal Age............................. 49
Prenatal Care............................ 50
Education............................... 50
Legitimacy............................... 51
VI. DISCUSSION OF RESULTS........................ 52
VII. CONCLUSIONS.......................... 58
BIBLIOGRAPHY........................................ 60
APPENDIX
A. RAW DATA..................................... 64
B. MAP OF COLORADO COUNTIES AND ALTITUDES..
68


Vll
TABLES
Table
1. Mortality Rates and Percentage of Low-
Birth Weight Births by Origin: Colorado
1979-1982.............................. 14
2. Neonatal Mortality Rates for Selected
States by Altitude of Residence: 1950-
1980........................................ 18
3. Infant Mortality Rates for Selected
States by Altitude of Residence: 1950-
1980....................................... 19
4. Mean Gestation by Altitude and Sex:
Colorado 1979-1982............................ 29
5. Birth Weight and Infant Mortality Rates
by Altitude and Term: Colorado 1979-
1982.......................................... 30
6. Relationships Between Altitude, Birth
Weight and Mortality: Colorado 1,979-
1982....................................... 34
7. Relationships Between Altitude, Birth
Weight, Mortality and Sex: Colorado
1979-1982..................................... 36
8. Mortality Rates by Birth Weight and
Altitude: Colorado 1979-1982.................. 37
9. Birth Weight by Complications: Colorado
1979-1982................................... 43
10. Altitude by Complications: Colorado
1979-1982..................................... 43
11. Leading Causes of Death by Altitude:
Colorado 1979-1982............................ 45
12. Birth Weight by Mother's State of Birth
and Altitude: Colorado 1979-1982
46


VI11
TABLES (Continued)
Table
13. Percentages of Low Birth Weight Births
by Socioeconomic Variables: Colorado
1979-1982................................ 47
14. Variation in Socioeconomic Variables by
Altitudes Colorado 1979-1982............. 48


ix
FIGURES
Figure
1. Incidence of Low Birth Weight in the
United States and Colorado 1950-1980.... 20
2. Percentages of Low Birth Weight Births
by Altitude: Colorado 1969-1972 and
1979-1982..................................... 22
3. Infant Mortality Rates by Altitude:
Colorado 1969-1972 and 1979-1982............ 23
4. % Preterm Births and % Preterm Deaths
by Altitude: Colorado 1969-1972 and
1979-1982..................................... 24
5. X Birth Weight by Altitude, Total
Births: Colorado 1979-1982.................. 25
6. X Birth Weight of Neonatal Deaths by
Altitude: Colorado 1979-1982................ 26
7. X Birth Weight of Infant Mortalities by
Altitude: Colorado 1979-1982................ 27
8. Neonatal Mortality Rates by Altitude:
Colorado 1979-1982............................ 32
9. Infant Mortality Rates by Altitude:
Colorado 1979-1982............................ 33
10. X Birth Weight by Gestational Age by
Altitude: Colorado 1979-1982............... .38
11. % Births in Level III Institutions by
Altitude: Colorado 1979-1982............... 40
12. % Low Birth Weight Births in Level III
Institutions by Altitude: Colorado
1979-1982..............................
42


CHAPTER I
INTRODUCTION
Low birth weight is universally associated with
increased probability of infant death. Approximately
7% of the infants born in the United States weigh
2500g or less (McCormick 1985). This 1% accounts for
almost two thirds of the neonatal deaths in this coun-
try (McCormick 1985).
Numerous studies conducted over the past 30
years indicate that there is a causal relationship be-
tween high altitude (elevations above 2500M) and low
birth weight (<2500g) (Lichty et al. 1957; Grahn and
Kratchman 1963; McClung 1969; Weinstein and Haas 1977;
McCullough, Reeves and Liljegren 1977; Beall 1981).
The precise mechanism responsible for this increased
incidence of low birth weight at high altitude is not
fully understood, but is believed to result from
stresses imposed by the high altitude environment. The
delineation of the exact etiology of altitude-associated
reductions in birth weight and their effects on infant
mortality is relevant not only to the study of differing
human physiological and genetic adaptations to environ-
mental stress, but also to the prevention of low birth


2
weight and improvement of the care and prognosis for
low birth weight babies born at any altitude.


I
CHAPTER II
BACKGROUND
Human Adaptation
Life at high altitude imposes a unique combin-
ation of environmental stresses, including decreased
oxygen availability; cold, arid and windy climate;
high levels of solar radiation; rugged terrain; and
shortened growing season. Recognition of the diffi-
culties of life in the high altitude environment ex-
isted when the Spanish colonized the Andes in the
17th century. Spanish historians reported that both
colonists and their imported domestic animals had great
difficulty reproducing in the high altitude regions of
the New World (Moore and Regensteiner 1980). Not only
do previous reports indicate that birth weights de-
crease and infant mortality rates increase at high al-
titude, but pregnant mothers have a greater incidence
of pre-eclampsia, and neonates have significantly
higher rates of hyperbilirubinemia (Moore and Regen-
steiner 1983). As the greatest postnatal mortality
risk occurs during the first year of life, the exist-
ence of more pregnancy-related and neonatal problems


4
at high altitude implies that human adaptation to high
altitude is an incomplete process, although populational
differences probably exist in different regions, depend-
ing upon genetic make-up and length of residence at
high altitude.
A decrease in oxygen availability, known as
hypoxia, is the primary challenge of the high altitude
environment because its effects cannot be counteracted
by cultural means. A constant supply of oxygen is
necessary to support human life. Because there is less
atmospheric pressure at high altitude, the partial
pressure of oxygen is reduced. Therefore, an indivi-
dual must adapt physiologically in order to maintain the
necessary flow of oxygen to the body tissues (Mazess
1975, Frisancho 1979, Moore et al. 1980). These adapta-
tions include greater lung capacity and efficiency, in-
creased hematocrit and reduced affinity of hemoglobin
for oxygen. Whether the mechanisms underlying these
adaptations are genetic, acquired, or some combination
of the two, is not known at the present time.
The combination of the chronic hypoxia exper-
ienced by the mother during pregnancy at high altitude,
the fact that oxygen deprivation can seriously affect
physiological functioning, and the fact the the fetus
develops in an already hypoxic environment, has resulted
in hypoxia being implicated as the most likely cause of


5
the reductions in birth weight at high altitude (Grahn
and Kratchman 1963; McCullough, Reeves and Liljegren
1977; Moore et al. 1982). Researchers hypothesize that
exaggerated fetal hypoxia at high altitude may reduce
the metabolism of nutrients necessary to sustain the
extremely rapid fetal growth, thus resulting in lower
birth weights (Moore and Regensteiner 1983:292). Rein-
\
forcing this hypothesis is research done by Moore et al.
(1982) which indicates that infant birth weight is re-
lated to maternal arterial oxygenation at high altitude.
Birth Weight as an Indicator of Survivorship
Birth weight, interpreted as an easily measured
reflection of developmental maturity, has long been
recognized as the most obvious predictor of infant sur-
vival during the first year of life (McCormick 1985:83).
Because birth weight is related to length of gestation,
these two variables must be considered together in order
to accurately assess the relationship between birth
weight and infant mortality. For any given birth
weight, infants whose weight is appropriate for their
gestational age (AGA) and those who are small for their
gestational age (SGA) differ in terms of the etiology
and effects of the low birth weight (McCormick 1985:
84). Results of studies on the relative contributions
of babies who are small for gestational age versus in-


6
fants who are appropriate for gestational age to infant
and neonatal mortality are inconclusive. A Colorado
study of births occurring between 1974 and 1980 (Koops,
Morgan and Battaglia 1982:971) indicated that infants
who were SGA had consistently higher neonatal mortality
rates than those who were AGA. A previous study done
in the 1960s by van den Berg and Yerushalmy (in
McCormick 1985:84) showed premature infants who were
SGA having lower neonatal and infant mortality rates
than infants who were AGA.
In 1980, infant mortality in the United States
was 13.1 deaths per 1,000 live births, down 47% from
the 1965 rates (McCormick 1985). McCormick attributes
this reduction to the increased survival of low birth
weight babies, resulting from more intensive hospital-
based management, since the incidence of low birth
weight declined only .7% between 1950 and 1980.
Birth Weight and Infant Mortality at High Altitude
Although all of the previous studies agree on
the correlation between increasing altitude and de-
creasing birth weight, interpretations of the effects
of the decreased birth weight on mortality at high al-
titude have differed. Lichty et al. (1957) determined
that neonatal mortality and the incidence of low birth
weight were significantly higher from 1949-1953 in Lake


7
County, Colorado's highest county, than in the rest of
the state. Grahn and Kratchman (1963) found that neo-
natal mortality rates and the frequency of low birth
weight were greater in most of the mountain states dur-
ing the 1950-57 period than in the rest of the country.
Later studies indicate a much weaker relation-
ship between low birth weight at high altitude and in- ,
creased neonatal and infant mortality. Frisancho and
Cossman (1970) compared birth weights and mortality
rates for low and high altitude states during the
period 1957-1967. They found that the incidence of low
birth weight in the high altitude states remained
fairly constant over the 11 year period. The neonatal
mortality rate, however, declined an average of 4
deaths per 1,000 live births in the high altitude
states, to a level equivalent to the low altitude rates.
They hypothesized that the increased neonatal mortality
which Grahn and Kratchman reported during the early
1950s was probably the result of limited medical facil-
ities and poorer socioeconomic conditions in the high
altitude states, rather than the effects of high alti-
tude hypoxia. McCullough, Reeves and Liljegren (1977)
studied Colorado vital statistics from 1960-1973, com-
paring neonatal and infant mortality rates for counties
grouped according to altitude. They discovered that,
"Increased infant mortality at the higher altitudes in


8
Colorado was primarily related to shortened gestation-
al time, rather than to low birth weight per se
(1977:38)." Beall (1981) in her analysis of birth
weight-specific mortality rates in a high altitude and
in a low altitude Peruvian community, found that the
high altitude population had a lower optimum birth
weight and a lower incidence of birth weight associated
mortality than the low altitude group. Thus, the re-
sults of the more recent studies correlating altitude,
birth weight and mortality indicate that the occurrence
of low birth weight in a high altitude environment may
not be as closely related to mortality as previously
believed.


9
CHAPTER III
RESEARCH OBJECTIVES
The objective of the present study was to find
the contribution of birth weight to infant mortality
at high altitude in the State of Colorado. Analyzing
all live births and infant and neonatal deaths occur-
ring from 1979-1982, we sought to determine:
1. The association between altitude and
birth weight.
2. The association between altitude and
infant mortality.
3. Whether the relationships between birth
weight, gestational age and mortality
were the same at all altitudes, or whether
they varied in relation to altitude.
4. Whether the cause of death, complications
of pregnancy, complications of labor,
and congenital malformations had any re-
lation to altitude.
5. Whether other maternal characteristics
which relate to low birth weight and/or
infant mortality varied with altitude.


CHAPTER IV
METHODOLOGY
Site Selection
The State of Colorado was selected as the study
location because Colorado is the only high altitude re-
gion in the world which possesses substantial popula-
tion numbers residing over a broad altitude range
(3,000'-10,200'+) and good quality birth and death
records for the entire' population. In addition, a
number of previous studies (Lichty et al. 1957; Grahn
and Kratchman 1963; Weinstein and Haas 1977; McCullough
Reeves and Liljegren 1977; Moore et al. 1982) have
been conducted using Colorado data, so that histori-
cal comparisons were possible.
The Colorado high altitude population differs
from African, Andean and Himalayan populations in
terms of the short duration of residence at high alti-
tuce. This shorter residence period eliminates the
possibility of the existence of long-term population
adaptations to high altitude. Conversely, the short
residence period may be advantageous because the ef-
fects of high altitude may be more readily apparent in
individuals who have recently migrated to high alti-


11
tude, with little time to adapt genetically.
Data
Data utilized in this study were derived from
computerized birth and death certificates of 186,616
surviving births and 1,684 infant deaths which occurred
between 1979 and 1982 in Colorado. This information
was compiled by the Health Statistics Section of the
Colorado State Department of Health. The University
of Colorado/Denver computing facilities were used to
analyze a tape of the combined birth and death certi-
ficates. Due to Colorado State Department of Health
policies designed to protect confidentiality, the tape
of the birth certificates of the surviving babies could
not be obtained. Data on surviving births came from
computer analyses routinely published by the Health
Department and a number of computer runs which the
Health Statistics Section completed especially for
this project.
The accuracy of the data is, of course, depend-
ent upon the accuracy of the information originally
obtained from the mother and recorded on the birth and
death certificates, and also upon the accuracy of the
coding of the data for input into the computer. Items
judged as mistakes in the data were recorded as missing
values.


12
Definitions
An infant death is defined as any death occur-
ring between birth and one year of age. Any infant
born in 1982 who died prior to one year of age was in-
cluded in this study. Neonatal deaths are defined as
deaths occurring prior to 28 days of age. Although
many previous studies have tabulated information on
neonatal deaths only, prenatal influences could well
affect the baby's survival throughout the entire cru-
cial first year of life. Therefore, both neonatal and
infant deaths were included in this report. All mor-
tality rates have been computed as the number of deaths
per 1,000 live births.
All Colorado resident births occurring between
1979 and 1982 were included in this study. The alti-
tude assigned to each birth and death was the mother's
county of residence stated on the birth certificate.
The altitude figures for each county are based on the
mean population elevation as calculated by the Colorado
State Department of Health. The counties were grouped
on the basis of altitude intervals of 3,000'-4,999'
(N=31,840 births), 5,000'-6,999' (N=143,240 births),
7,000'8,999' (N=9,453 births) and 9,000'+ (N=2,083
births).


13
According to McCormick (1985:83), black infants
are twice as likely to have low birth weight values as
white infants (12.7% versus 5.9%, respectively). In
this Colorado sample, the mean birth weight for the
white babies was 2028g. The mean birth weight for the
black babies was 1668g. Because the proportion of
black to white births in this sample was not constant
at each altitude level, only white births were included
in statistics tabulated for this paper, unless otherwise
noted.
The "White" classification includes Colorado's
Hispanic population. The birth certificate has two
categoriesSpanish Surname and Originto determine
Hispanic descent. Hispanic classification on the basis
of Spanish Surname can include individuals whose ori-
gins are as diverse as Spain, South America, Mexico,
Puerto Rico, etc. In addition, the names classified in
\
the Spanish Surname category by the Health Department
vary from year to year. Classification of Hispanics
using the Origin category is not possible because
many Hispanics born in the United States will state
their origin as American. Mortality and birth weight
statistics for Colorado's Hispanic population are
tabulated in Table 1 on the basis of the Spanish
Surname category. Because the determination of exact
Hispanic origin is so uncertain, and because the rates


TABLE 1
MORTALITY RATES* AND PERCENTAGE OF LOW BIRTH WEIGHT BIRTHS BY ORIGIN
COLORADO 1979-1982
ORIGIN MEAN BIRTH WEIGHT NEONATAL MORTALITY RATE INFANT MORTALITY RATE % LOW BIRTH WEIGHT BIRTHS MORTALITY RATE OF LOW BIRTH WEIGHT BABIES
ANGLO 2028g 5.5 00 00 .74% 6.4
HISPANIC** 1999g 6.1 9.4 .91% 6.0
BLACK 1668g 8.3 13.5 1.33% 9.8
Deaths per 1,000 live births
**As determined by Spanish Surname


15
in Table 1 are so similar to the Anglo rates, Hispanic
statistics were not separated from the Anglo statsitics
in any further analyses.
Statistics
The significance level for all statistics com-
puted in this paper is .05. Missing values have been
excluded from all analyses. The chi-square test and
regression analysis have been used, as noted, to
determine whether statistically significant differences
exist among the variables analyzed.
2
The chi-square (X ) test requiring few assump-
tions and only nominal level data, is used to evaluate
whether or not measured values differ significantly from
those which would be expected under a specific set of
theoretical assumptions (Blaylock 1979:279). The sig-
nificance of, a chi-square value is dependent not only
upon the degree of the relationship between the vari-
ables, but also upon the size of the sample. With a
very large sample, such as was utilized in this study,
statistical significance can be easily established,
even when there is a very slight relationship between
variables (Blaylock 1979:300). Therefore, Cramer's V,
a measure of the strength of association, was computed
as part of each statistical analysis. The value of
Cramer's V (C ) ranges from zero to one, with zero in-


16
dicating no relationship between the variables, and a
value of one indicating a perfect relationship between
the variables (Blaylock 1979:305).
In certain situations comparing the relation-
ship between two interval level variables, regression
equations were used to determine statistical signifi-
cance. The regression equation is actually a plot of
the means of the dependent variable for every fixed
value of the independent variable. The actual regres-
sion line is drawn in the area where the sum of the
squares of the deviations of the actual values from
the line are at a minimum (Blaylock 1979:391). To
determine the strength of this relationship, Pearson's
r was calculated. This coefficient of correlation
measures the amount of spread about the linear least
squares equation (Blaylock 1979:396).. The range for r
is -1.0 to 1.0. If all points fall exactly on the
regression line, r will equal either 1.0 or -1.0.,
depending on whether the relationship is positive or
negative (Blaylock 1979:397).


CHAPTER V
RESULTS
Historical Perspective
Tables 2 and 3 compare Colorado's neonatal and in-
fant mortality rates from 1950-1980 with rates for the
entire United States, all of the other high altitude
states, and a selected number of low elevation states.
The total decline in infant mortality in Colorado during
this period was 24.1 deaths per 1,000 live births, as
compared to an overall decline of 17.5 deaths per 1,000
births in the United States as a whole. Similar large
declines have also occurred in the other mountain
states. In 1980, Colorado's infant mortality rate
(10.1) was the fifth lowest in the country. Only
Maine (9.2), Wyoming (9.8), New Hampshire (9.9), and
Minnesota (10.0) had lower rates.
Figure 1 compares the incidence of low birth
weight in the United States for selected years from
1950 through 1980 to Colorado rates. Over this 30 ,year
period, the national low birth weight rate has decreased
only .7%, as compared to a decrease in Colorado of 2.7%.
The relatively greater decrease in Colorado occurred
primarily between 1950 and 1955.


TABLE.2
NEONATAL MORTALITY RATES* FOR SELECTED STATES BY ALTITUDE OF RESIDENCE
X ALTITUDE: 5390' 5390 5011' 4590' 3690* 3490' 3320* 2080' 1440' 630' 173'
YEAR U.S. COLO. WYOMING NEW MEXICO UTAH NEVADA MONTANA IDAHO ARIZ. WASH. INDIANA FLORIDA
1950 20.5 23.4 24.1 26.6 17.3 27.8 19.0 19.8 25.6 19.9 18.7 15.0
1955 19.1 22.4 21.1 24.0 15.9 23.6 18.6 15.9 20.1 17.5 18.7 20.9
1960 18.7 21.3 21.5 20.2 14.5 21.0 17.1 17.3 19.3 17.2 17.5 20.8
1965 17.7 17.7 16.3 17.7 13.5 19.4 18.1 17.4 15.9 .15.2 17.5 19.8
1970 15.1 15.1 15.4 14.8 11.1 18.2 15.9 13.1 12.4 13.4 14.9 16.1
1975 11.6 9.5 12.8 10.9 9.3 11.6 10.6 9.4 9.6 10.9 10.6 12.5
1976 10.9 8.7 10.8 9.7 7.9 10.7 11.7 8.3 10.4 9.3 10.4 10.7
1977 9.9 7.4 8.3 9.6 6.8 8.9 8.9 6.7 8.7 . 7.4 9.7 10.6
1978 9.5 6.9 8.1 9.0 7.8 8.6 7.3 7.3 9.0 7.9 8.6 9.3
1979 8.9 6.4 8.4 8.9 6.8 6.6 7.1 6.6 8.9 7.1 9.2 9.9
1980 8.4 6.6 6.2 7.3 6.6 6.7 7.7 6.2 8.3 7.0 7.9 9.8
Source: U.S. Department of Health, Education and Welfare, Public Health Service,
National Office of Vital Statistics, Vital Statistics of the United States, 1955, 1960, 1965,
1970, 1975, 1980 (Washington, D.C.: U.S. Government Printing Office, 1957, 1963, 1968, 1975,
1978, 1984).
Deaths per 1,000 live births
oo


TABLE 3
INFANT MORTALITY RATES* FOR SELECTED STATES BY ALTITUDE OF RESIDENCE
X ALTITUDE: 5390 5390' 5011' 4590' 3690' 3490' 3320' 2080' 1440' 630' 173'
YEAR U.S. COLO. WYOMING NEW MEXICO UTAH NEVADA MONTANA IDAHO ARI2. WASH. INDIANA FLORIDA
1950 29.2 34.2 32.5 54.8 24.2 37.9 28.2 27.1 45.8 27.3 27.0 19.0
1955 26.4 30.2 28.5 43.0 20.5 30.0 25.0 20.8 34.8 24.5 25.0 29.7
1960 26.0 27.5 28.2 35.5 19.6 30.1 25.0 22.9 31.9 23.4 23.9 29.7
1965 24.7 24.3 22.1 26.9 18.7 24.8 25.0 24.5 25.4 21.4 23.6 28.3
1970 20.0 20.0 20.0 21.0 14.9 24.7 21.5 17.1 17.7 18.7 19.4 21.5
1975 16.1 14.2 17.4 17.0 13.1 17.4 15.4 13.2 14.8 15.8 14.8 17.7
1976 15.2 13.0 16.5 15.5 11.7 14.2 16.5 13.1 15.4 14.5 14.6 15.1
1977 14.1 12.0 13.9 13.8 10.1 13.9 13.7 11.3 13.5 12.1 14. 2 15.5
1978 13.8 11.2 13.0 14.1 11.4 12.5 11.6 11.7 13.1 12.5 13.1 14.1
1979 13.1 10.6 13.2 14.0 10.7 10.7 10.7 10.0 13.6 11.5 13.0 14.9
1980 ,12.5 10.1 9.8 11.5 10.4 10.7 12.4 10.7 12.4 11.8 11.9 14.6
Source: U.S. Department of Health, Education and Welfare, Public Health Service,
National Office of Vital Statistics, Vital Statistics of the United States, 1955, 1960, 1965,
1970, 1975, 1980, (Washington,' D.C.: U.S. Government Ptintihg Office, 1957, 1963, 1968, 1975,
1978, 1984).
Deaths per 1,000 live births


20
% LOW BIRTH
WEIGHT BIRTHS
COLORADO
UNITED STATES
FIGURE 1
INCIDENCE OF LOW BIRTH WEIGHT IN THE
UNITED STATES AND COLORADO 1950-1980


21
Figures 2 and 3 compare the percentages of low
birth weight babies and infant mortality rates in Colo-
rado for infants born between 1969 and 1972 to the
1979-1982 sample analyzed in this study. These fig-
ures clearly indicate that the greatest declines in
both frequency of low birth weight and infant mortality
have occurred at high altitude, with decreases above
9,000' over twice as great as those below 5,000'.
Comparing the two figures, the decreases in infant
mortality over the past 10 years have been consider-
ably greater than the decreases in incidence of low
birth weight.
Figure 4 compares the current incidence of
preterm births and preterm mortality rates to the
1969-1972 sample of Colorado births analyzed by
McCullough, Reeves and Liljegren (1977). Incidence of
preterm births has declined over fourfold during the
past 10 years. Preterm infant mortality rates, how-
ever, are generally higher for the current sample,
except at the highest altitude level.
Relationships Between Altitude, Birth Weight
and Gestation
Figures 5, 6, and 7 illustrate the relation-
ships between altitude and mean birth weight by sex,
for total births, neonatal mortalities and infant mor-
talities. The results of a linear regression comparing


% LOW BIRTH
WEIGHT BIRTHS
22
ALTITUDE: 3-4999' 5-6999' 7-8999' 9000'+
FIGURE 2
PERCENTAGES OF LOW BIRTH WEIGHT BIRTHS BY ALTITUDE
COLORADO 1969-1972 and 1979-1982


23
INFANT MORTALITY RATE
(Deaths per 1000 live births)
1969-1972
FIGURE 3
INFANT MORTALITY RATES BY ALTITUDE
COLORADO 1969-1972 and 1979-1982


% PRETERM
38 Weeks)
BIRTHS
DEATHS/100
PRETERM BIRTHS
ALTITUDE: <7000' 7-8999' 9000'+
FIGURE 4
% PRETERM BIRTHS AND % PRETERM DEATHS BY ALTITUDE
COLORADO 1969-1972 and 1979-1982
N)


25
X BIRTH WEIGHT
FIGURE 5
X BIRTH WEIGHT OF TOTAL BIRTHS BY ALTITUDE
COLORADO 1979-1982


X BIRTH WEIGHT TOTAL POPULATION
X BIRTH WEIGHT BY SEX
1800g
1700g
1600g
1500g
1400g
1300g
ALTITUDE:
FIGURE 6
X,BIRTH WEIGHT OF NEONATAL DEATHS BY ALTITUDE
COLORADO 1979-1982
to
OI


27
X BIRTH WEIGHT
2300g
2200g
2I00g
2000g
1900g
1800g
1700g
1600g
1500g
ALTITUDE
FIGURE 7
X BIRTH WEIGHT OF INFANT MORTALITIES BY ALTITUDE
COLORADO 1979-1982


28
the mean birth weights of the total births by mean al-
titude values indicate that there is a statistically
significant inverse relationship between altitude and
birth weight. The coefficient of correlation
(Pearson's r) for this relationship is -.98. There
appears to be no significant relationship between the
birth weights of either the neonatal or the infant non-
survivors and altitude. Surviving males average 133.3g
heavier than females. Among the deaths, the males
are consistently heavier than the females, except at
the highest altitude level, where female neonatal mor-
talities average 290g heavier than males, and female
infant mortalities average 387g heavier than males.
Table 4 indicates that there are no significant
differences in mean gestational age by either altitude
or sex, so the lower female and high altitude birth
weights are not the result of increased numbers of
preterm births. When birth weight is tabulated by
term in Table 5, the relationships between birth weight
and altitude are not altered. After excluding all
preterm birth weights, the greatest incidence of
low birth weight among full-term babies occurs above
9,000 feet.


TABLE 4
MEAN GESTATIONAL AGE BY ALTITUDE AND SEX
COLORADO 1979-1982
MEAN GESTATIONAL AGE MEAN GESTATIONAL AGE MEAN GESTATIONAL AGE
ALTITUDE TOTAL BIRTHS FEMALES MALES
000' ' + 39.5 Weeks 39.9 Weeks 39.8 Weeks
000' 1 -8,999' 39.9 Weeks 39.9 Weeks 39.9 Weeks
000' '-6,999' 40.0 Weeks 40.1 Weeks 40.0 Weeks
3,000' -4,999'
40.1 Weeks
40.1 Weeks
40.0 Weeks


TABLE 5
X BIRTH WEIGHTS AND INFANT MORTALITY RATES* BY ALTITUDE AND TERM
COLORADO 1979-1982
TERM BIRTHS PRETERM BIRTHS**
ALTITUDE: 3-4,999' 5-6.999' 7-8,999' 9,000'+ 3-4,999' 5-6,999' 7-8,999' 9,000'+
TOTAL BIRTHS 29,350 131,441 8,606 1,925 1,493 6,294 418 86
% TERM/PRETERM 95.2 95.4 95.4 95.7 4.8 4.6 4.6 4.3
TOTAL SURVIVORS 29,208 130,831 8,555 1,916 1,372 5,727 380 80
TOTAL DEATHS 142 618 51 7 121 567 38 6
TOTAL LOW BIRTH WEIGHT BIRTHS 1,344 6,657 570 204
% LOW BIRTH WEIGHT 4.6 5.1 6.6 10.6
TOTAL LOW BIRTH WEIGHT SURVIVORS 1,307 6,522 554 197
TOTAL LOW BIRTH WEIGHT DEATHS 37 135 16 3
MORTALITY RATE TOTAL BIRTHS 4.8 4.7 5.9 4.7 81.0 90.1 90.9 69.8
MORTALITY RATE TOTAL LOW BIRTH WEIGHT BIRTHS 27.5 20.3 28.1 14.7
X BIRTH WEIGHT TOTAL BIRTHS 3,317g 3,286g 3,204g 3,105g 2,429g 2,285g 2,125g 2,125g
X BIRTH WEIGHT SURVIVORS Standard Deviation 3,319g 488 3,287g 481 3,207g 467 3,108g 468 2,540g 765 2,399g 754 2,354g 719 2,231g 718
X BIRTH WEIGHT INFANT DEATHS Standard Deviation 2,867g 816 2,913g 689 2,714g 568 2,431g 671 1,165g 715 1,128g 727 1# 251g 730 709g 376
Deaths per 1,000 live births
Less than 38 weeks gestation OJ
O


31
Relationships Between Altitude/ Mortality
and Gestation
Figures 8 and 9 detail both neonatal and infant
mortality rates by altitude and sex. A chi-square test
indicates that there are no statistically significant
differences in either neonatal or infant mortality
rates by altitude. Male infants exhibit elevated mor-
tality rates at all altitudes. Females born above
9,000' have significantly lower mortality rates than
females born at any other elevation. Referring back to
Table 5, the preterm infant mortality rate is lowest at
the highest altitude level, while term infant mortality
rates show little variation by altitude.
Relationships Between Altitude, Birth Weight,
Mortality and Gestation
Table 6 illustrates the relationships between
altitude, birth weight and mortality. Contrary to ex-
pectation, the percentages of low birth weight babies
who died, both as neonates and infants, decrease as al-
titude increases. The percentages of both normal and
very low birth weight deaths do not vary significantly
by altitude, indicating that any survival advantage
which low birth weight babies might have at high alti-
tude must occur within the 1500-2500g weight range.
The figures in Table 5 indicate that this low birth
weight advantage is valid for both term and preterm


32
NEONATAL MORTALITY RATES
(Deaths per 1,000 live births)
10
7
0
ALTITUDE:
3-4,999' 5-6,999' 7-8,999' 9,000'+
FIGURE 8
NEONATAL MORTALITY RATES BY ALTITUDE
COLORADO 1979-1982


33
INFANT MORTALITY RATES
(Deaths per 1,000 live births)
12
11
10
9
8
7
6
5
f'
4
3
2
1
ALTITUDE: 3-4,999' 5-6,999' 7-8,999' 9,000'+
FIGURE 9
INFANT MORTALITY RATES BY ALTITUDE
COLORADO 1979-1982


TABLE 6
RELATIONSHIPS BETWEEN ALTITUDE. BIRTH WEIGHT AND MORTALITY
COLORADO 1979-1982
ALTITUDE # INFANT DEATHS/1,000 VERY LOW BIRTH # INFANT DEATHS/1,000 LOW BIRTH # INFANT DEATHS/1.000 NORMAL BIRTH # NEONATAL DEATHS/1.000 VERY LOW BIRTH # NEONATAL DEATHS/1,000 LOW BIRTH # NEONATAL DEATHS/1,000 NORMAL BIRTH
WEIGHT BIRTHS WEIGHT BIRTHS WEIGHT BIRTHS WEIGHT BIRTHS WEIGHT BIRTHS WEIGHT BIRTHS
9,000'+ 409 41 3.9 364 33 2.2
N=9 N=11 N=7 N=8 N=9 N=4
7,0001-8,9991 341 62 4.4 293 47 1.5
N=28 N=54 N=38 N=24 N=41 N=13
5,0001-6,9991 373 68 4.0 343 57 1.5
N=531 N=754 N=535 N=488 N=632 N=196
3,0001-4,999' 412 ( 78 4.0 385 62 1.9
N=108 N=166 N=119 N=101 N=136 N=55
COLORADO 378 70 4.1 347 58 1.6
N=676 N=985 N=699 N=621 N=818 N=268


35
births.
Table 7 details the relationships between alti-
tude, birth weight and mortality by sex. The data in-
dicate that the survival differential between males
and females of normal birth weight is very slight.
Among the low birth weight population, female mortality
is consistently lower than male mortality, with the
greatest differential occurring at the highest altitude
Among the very low birth weight group, mortality seems '
to be less related to sex, with females from the 5,000'-
6,999' level experiencing greater mortality than males.
t
At high altitude, however, females again have the great-
est advantage over males. Although there are fewer
males born with low birth weights, the proportion of
male low birth weight babies who die as infants is
greater than the proportion of female low birth weight
babies who die as infants. When the mortality rates
are tabulated by sex, the reduced mortality rates for
low birth weight babies at high altitude appear to be
valid for all females weighing less than 2500g and
for males weighing between 1500 and 2500g.
Table 8 shows the distribution of mortality
rates by altitude and birth weight. Babies weighing
between 3501 and 4500g have the lowest mortality rates
at all altitude levels.
In Figure 10, the mean birth weights by ges-


TABLE 7
RELATIONSHIPS BETWEEN ALTITUDE, BIRTH WEIGHT, MORTALITY AND SEX
COLORADO 1979-1982
ALTITUDE % FEMALE LOW BIRTH WEIGHT BIRTHS % MALE LOW BIRTH WEIGHT BIRTHS % FEMALE VERY LOW BIRTH WEIGHT BIRTHS % MALE VERY LOW BIRTH WEIGHT BIRTHS % NORMAL BIRTH WEIGHT FEMALE INFANT DEATHS % NORMAL BIRTH WEIGHT MALE INFANT DEATHS % LOW BIRTH WEIGHT FEMALE INFANT DEATHS % LOW BIRTH WEIGHT MALE INFANT DEATHS
9,000'+ 14.3% 11.7% 0.95% 1.2% 0.33% 0.44% 2.0% 6.7%
N=151 N=120 N=10 N=12 N=3. N=4 N=3 N=8
7-8,999' 10.4% 8.1% 1.00% 0.7% 0.44% 0.45% 5.2% 7.3%
t N=478 N=395 N=48 N=34 N=18 N=20 N=25 N=29
5-6,999' 8.6% 6.9% 0.96% 1.1% 0.35% 0.45% 7.5% 8.6%
N=6008 N=5031 N=675 N=748 N=225 N=310 N=321 N=433
3-4,999' 7.6% 6.2% 0.85% 0.8% 0.24% 0.54% 7.3% 8.0%
N=1186 N=1005 N=132 N=130 N=35 N=83 N=86 N=80
COLORADO 8.7% 6.9% 0.95% 1.0% 0.34% 0.47% 7.1% 8.4%
N=7823 N=6551 N=865 N=924 N=281 N=417 N=435 N=550
W
CT.


TABLE 8
MORTALITY RATES* BY BIRTH WEIGHT AND ALTITUDE
COLORADO 1979-1982
ALTITUDE
BIRTH WEIGHT 3-4,999' 5-6,999' 7-8,999' 9,000'+
>5,001g 62.5 . 34.5 0 0
4,501-5,OOOg 10.0 3.0 0 0
4,001-4,500g 2.6 2.8 3.0 0
3,501-4,OOOg 2.8 2.7 1.0 0
3,001-3,500g 3.8 3.7 3.1 2.4
2,501-3,OOOg 6.0 6.6 9.5 8.3
2,001-2,500g 21.0 16.6 25.0 9.2
1,501-2,OOOg 69.3 50.0 66.7 0
1,001-1,500g 197.1 172.4 227.3 111.1
501-1,OOOg 603.8 610.1 722.2 545.4
< 500g 894.7 853.3 833.3 1000.0
*Deaths per 1,000 live births


38
X BIRTH WEIGHT
(weeks)
FIGURE 10
X BIRTH WEIGHT BY GESTATIONAL AGE BY ALTITUDE
COLORADO 1979-1982


39
tational age for each altitude category are illustrated.
Babies born below 5,000' are clearly heavier at every
gestation interval. Birth weight values at high alti-
tude are greater than or equal to birth weight values
at the other altitude levels until week 30. Beyond
this point, the graph assumes the expected configura-
tion, with birth weight inversely related to altitude.
Level of Care
The overwhelming majority of all infants, re-
gardless of altitude, were born in hospitals, and the
births were attended by physicians. Although there is
no formal rating system, newborn nurseries are classi-
fied as Level I, Level II and Level III, according to
a model recommended by the Committee on Perinatal
Health (Paneth et al 1982:149). Level III neonatal
intensive care facilities provide the most advanced
care, and are generally operated as regional centers
which accept referrals from other hospitals in the
area. A study done by Paneth et al. (1982) indicated
that birth at a Level III center does lower neonatal
mortality in low birth weight infants. All of the
Level III facilities in Colorado are located in the
Denver metropolitan area. Figure 11 tabulates the
number of babies who were transported to Level III
facilities in the Denver area for birth. These
figures exclude births occurring in the 5-6,999' alti-


40
% BIRTHS
IN LEVEL III
INSTITUTIONS
FIGURE 11
% BIRTHS IN LEVEL III INSTITUTIONS BY ALTITUDE
COLORADO 1979-1982


41
tude interval. Since all of the Level III institutions
are located within this elevation category, births
from this altitude level occurrring in these facilities
are less likely to reflect any intentional choice re-
garding care availability as compared to families who
have traveled from other altitude levels to give birth
in Level III institutions. A chi-square test indicates
that significantly more babies from the highest alti-
tude level were born in Level III facilities, although
the relationship is quite weak (Cv=.0009). When the
low birth weight babies are tabulated by institution
type and altitude in Figure 12, slightly more births
from the 7-8,999' level were born in Level III facili-
ties. Mortality rates for both low and very low birth
weight babies were the lowest in this altitude category.
Complications
Tables 9 and 10 detail complications of preg-
nancy, complications of labor and congenital malforma-
tions by birth weight and altitude. The low birth
weight babies experience significantly more problems
than the normal weight babies. With the exception of
the congenital malformation category, chi-square tests
show complications of both pregnancy and labor having
a significant relationship to altitude, although the
coefficients of correlation are quite low.


42
% LOW BIRTH WEIGHT BIRTHS
IN LEVEL III INSTITUTIONS
FIGURE 12
% LOW BIRTH WEIGHT BIRTHS IN
LEVEL III INSTITUTIONS BY ALTITUDE
COLORADO 1979-1982


TABLE 9
BIRTH WEIGHT BY COMPLICATIONS
COLORADO 1979-1982
LOW BIRTH WEIGHT NORMAL BIRTH WEIGHT
COMPLICATIONS OF PREGNANCY, % 28.7 6.9
COMPLICATIONS OF LABOR, % 44.9 21.6
CONGENITAL MALFORMATIONS, % 2.4 0.9
TABLE 10
ALTITUDE BY COMPLICATIONS
COLORADO 1979-1982
COMP. OF COMP. OF CONGENITAL
ALTITUDE PREGNANCY LABOR MALFORMATIONS
3-4,999' 7.3% 20.1% 1.2%
5-6,999' aft CO CD 23.9% o i1
7-8,999' 9.6% . 25.1% 1.3%
9,000'+ C& o i1 il 27.9% 11


44
Cause of Death
The cause of death reported on the death certi-
ficate was checked in order to determine whether there
was any relations to altitude. Table 11 illustrates
the incidences of the most frequent causes of death by
altitude. Incidence of respiratory problems is lowest
at the highest altitude level, with heart and genetic
disorders more prevalent.
Mother's Altitude of Birth
Although the exact location of mother's birth
was not available from the data utilized for this
study, the data presented in Table 12 indicate that
Colorado native mothers tend to produce a slightly
higher percentage of low birth weight babies, regard-
less of altitude, than mothers born in other, presum-
ably lower, altitude states.
Socioeconomic Factors
Many socioeconomic factors, such as maternal
age, smoking, alcohol and drug consumption, prenatal
care, etc., are believed to influence birth weight.
Those variables for which information was available on
the birth certificates were analyzed. Table 13 corre-
lates the percentages of low birth weight babies with the
the socioeconomic variables. Table 14 details the fre-


TABLE 11
LEADING CAUSES OF DEATH BY ALTITUDE
COLORADO 1979-198 2
3,0001-4, 999' 5,000'-6, 999' 7,000'-8,999 ' 9,000'+
SIDS 18.3% Respiratory 19.0% SIDS 16.8% Heart 22.2%
Respiratory 15 2% SIDS 16.0% Respiratory '16.8% Sepsis 16.7%
Genetic 12.4% Genetic 12.8% Genetic 12.6% Genetic 16.7%
Heart 9.7% Heart 9.3% Heart 12.6% Respiratory 11.1%
Sepsis 7.6% Sepsis 8.4% Sepsis 11.6% SIDS 11.1%
Lack of Lack of Placental 5.3% Accident 5.6%
Information 20.7% Information 15.2% Lack of Lack of
Other 16.1% Other 19.3% Information 7.4% Information 16.7%
Other 16.9% Other 0.0%
.L
ui


TABLE 12
BIRTH WEIGHT BY MOTHER'S STATE OF BIRTH AND ALTITUDE
COLORADO 1979-1982
ALTITUDE % COLORADO NATIVE MOTHERS PERCENTAGES OF COLORADO NATIVES LOW BIRTH WEIGHT BABIES NON-COLORADO NATIVES
9,000'+ 37.3% 13.9% 12.5%
7-8,999' 42.8% 9.7% 8.8%
5-6,999' 34.3% 7.9% 6.8%
3-4,999' 54.0% 7.2% 6.6%
L
cn


47
TABLE 13
PERCENTAGES OF LOW BIRTH WEIGHT BIRTHS
BY SOCIOECONOMIC VARIABLES
COLORADO 1979-1982
SOCIOECONOMIC VARIABLE % LOW BIRTH WEIGHT BIRTHS
Maternal Age
< 20 Years 10.6%
20-39 Years 7.6%
>39 Years 10.4%
1 Trimester Prenatal Care Began
No Care 22.2%
3rd Trimester CO 00 a*
2nd Trimester 9.2%
3rd Trimester 9.3%
Mother's Education
<12 Years 10.9%
High School Graduates 8.2%
>12 Years 6.4%


48
TABLE 14
VARIATION IN SOCIOECONOMIC VARIABLES BY ALTITUDE
COLORADO 1979-1982
ALTITUDE
SOCIOECONOMIC VARIABLE 3' -4,999' 5-6,999* 7-8,999' 9,000'+
Population/Square Mile 14.6 52.5 4.4 6.0
Maternal Age
% <20 Years 15.5% 11.5% 11.9% 9.0%
% 20-39 Years 84.0% 88.0% 87.6% 90.8%
% >39 Years 0.5% 0.5% 0.5% 0.2%
Trimester Care Began
% No Care 0.6% 0.5% 0.5% 0.4%
% 3rd Trimester 5.0% 4.1% 4.6% 3.3%
% 2nd Trimester 20.0% 17.3% 19.8% 15.6%
% 1st Trimester 74.4% 78.0% 75.1% 80.7%
Mother's Education
% <12 Years 22.3% 17.3% 16.7% 11.4%
% High School Graduates 44.7% 41.5% 40.7% 41.0%
% >12 Years 33.1% 41.3% 42.6% 47.6%
% Illegitimate Births 14.6% 13.6% 6.6% 7.1%


49
quencies of these variables by altitude. The figures in
each of the following subsections represent the total
newborn population, with all races combined.
Urban vs. Rural. According to Taffel (1976:3),
mothers living in large urban places are more likely to
bear a low birth weight child than mothers residing in
less densely populated areas. The fact that the major-
ity of Colorado's urban population resides within the
5,000'-6,999' altitude category (See Table 14) indi-
cates that some factor other than an urban/rural differ-
ential is responsible for the increased incidence of
low birth weight at high altitude.
Maternal Age. Mother's age is related to both
infant mortality and birth weight, with either very
young or very old mothers likely to experience more
problems. According to Taffel (1976:2), nationwide in-
cidence of low birth weight in 1976 ranged from a high
of 14.8% for infants born to girls under 15 years of
age, to a low of 6%, for infants born to mothers in the
25-29 year old range. Mothers over 45 had a 10.5% in-
cidence of low birth weight babies. The figures for
Colorado mothers presented in Table 13 indicate that
mothers under 20 and over 40 have higher incidences of
low birth weight babies. Table 14 shows that there are
fewer mothers in the high risk age categories at the


I
50
highest altitude level.
Prenatal Care. The tabulation of trimester in
which care was begun, illustrated in Table 13, indi-
cates that the risk of low birth weight is only
exacerbated for cases in which the mother received no
prenatal care. The altitude figures in Table 14
show that babies born at altitude levels above 9,000'
and from 5,0011-7,0001 appear to have begun prenatal
care the earliest.
Education. Educational level is the closest
indicator of socioeconomic status available on the
birth certificate. The assumption is that education
and income are positively correlated and that regard-
less of income, more educated mothers are more con-
cerned and informed about prenatal and postnatal care
and nutrition than mothers who are less educated.
The figures in Table 13 show that there is a statis-
tically significant relationship between education
and incidence of low birth weight (C =.004), and the
figures in Table 14 indicate that mothers' educational
level increases with altitude.


51
Legitimacy. According to Taffel (1980:13),
babies born to unwed mothers tend to exhibit low birth
weights almost twice as often as babies born to
married mothers. The percentage of babies born to un-
married mothers at low altitude is more than double
the number born at high altitude (7.1% versus 14.6%).


CHAPTER VI
DISCUSSION OF RESULTS
, To explain the great declines in infant mor-
tality and low birth weight which have occurred in
Colorado over the past 10 years, a number of factors
must be considered. Declines in the incidence of low
birth weight are undoubtedly responsible for a propor-
tion of the decline in infant mortality. The huge
decline in incidence of preterm births must also have
affected both the mortality rates and incidence of low
birth weight. Overall improvements in medical care
and more widespread availability of this improved
medical care throughout Colorado may also be responsi-
ble for saving many lives. These theories, however, do
not explain why both preterm infant mortality and low
birth weight infant mortality rates are now lower at
high altitude than at lower altitudes.
Since the lower incidence of birth weight-
associated mortality at high altitude in Colorado
represents a change which has occurred only over the
past 10 years, Beall's interpretation of decreased
high altitude infant mortality resulting from differing
adaptational patterns among high and low altitude pop-


53
ulations (1981) cannot apply. A simpler explanation
is that the formerly high infant mortality rates at
high altitude resulted from inferior availability of
advanced medical care in the predominantly rural,
mountain areas of the state, rather than from the in-
creased incidence of low birth weight. One must assume
that medical facilities in the high altitude regions of
the state have improved considerably over the past 10
years, and that increased ability to transport endan-
gered infants to fully equipped neonatal intensive
care centers immediately after birth have been respon-
sible for a portion of the decreased infant mortality
rates at high altitude. The actual extent of care is
a variable which is almost impossible to determine
from the birth certificate data. An analysis of in-
dividual medical records would be necessary to verify
this hypothesis. Although the information in Figure 11
indicates that a number of high altitude parents do
travel to facilities offering Level III care for birth,
the fact that a baby was born in a hospital possessing
Level III facilities does not necessarily mean that
every infant born there utilized these facilities.
More low birth weight babies from the 7-9,000' altitude
category were born in Level III institutions, although
the low birth weight mortality rates for this group
were higher than for the babies born above 9,000'.


54
In addition, the number of low birth weight and/or
preterm infants born at high altitude and transported
immediately after birth to Children's Hospital' or some
other Level III institution is unknown.
Further research might prove that changes in
socioeconomic variables have taken place over the past
10 years which have affected prenatal and postnatal
care. The current available data indicate that high
altitude residents are comparatively advantaged in
terms of all socioeconomic indicators. Their status
10 years ago is unknown. These socioeconomic advan-
tages are undoubtedly increasing both birth weights
and survival rates at high altitude, -while simultane-
ously decreasing these values at lower elevations, thus
mitigating the full effects of altitude on birth weight
and mortality.
Preterm infant mortality rates have increased
over the past 10 years at every altitude level except
the 9,001'+ category. A possible explanation for this
increase in preterm mortality is that the few remain-
ing premature births which do occur happen very early
in the gestation period, thus having a much greater
mortality risk, or they are births accompanied by
by other serious problems.
Both surviving and non-surviving males consis-
tently weigh more than females, except at the highest


55
altitude level, where non-surviving females weigh sig-
nificantly more than non-surviving males. This infor-
mation, accompanied by the fact that the female surviv-
al advantage is greatest among the low birth weight
population at the highest altitude level, may indicate
that females are better adapted to survive at high al-
titude than males. There must be considerable reser-
vation, however, in generalizing about the birth
weights of infant mortalities at the highest altitude
level, as the sample size consisted of only 18 deaths
out of a total of 2,083 live births.
In a study conducted in Bolivia (1980), Haas
also concluded that females may be better adapted to
high altitude stress than males. His conclusion, how-
ever, was based on findings indicating that males were
consistently heavier than females at low altitude, but
this differential in birth weight disappeared at high
altitude.' The results of the present study do not
bear out his conclusion, as the greatest differential
between male and female birth weights occurs at the
highest altitude level in Colorado.
One question raised by his study and other
similar "high altitude" research projects is the defin-
ition of the exact elevation where altitude begins to
significantly affect adaptation. The results of this
analysis show a continuum of birth weight reductions
as altitude increases, with greater decreases at the


56
highest altitude levels. Historical changes, male
and female differences, and the percentages: of low
birth weight births appear to differentiate at alti-
tudes above 9,000'. If the inhabited elevations in
Colorado were extended to the 12,661' level of Beall's
(1981) Peruvian population, however, quite different
trends might appear. Another question involves the
validity of comparing populations who have been in-
digenous to high altitude for centuries, such as the
Andean groups, to populations such as the Colorado
group who have lived at high altitude for a compara-
tively short period.
The altitudes assigned to each birth in this
study were the mean population altitudes per county.
There is considerable altitude variation within certain
counties in the state. For example, babies gestated
in Nederland, Colorado, located in Boulder County,
actually lived at an altitude of 8,236', but were
classified at the 5,338' level assigned to Boulder
County. A more accurate method of analysis would have
been to classify altitude of birth by mother's city of
residence. Because the published Health Department
data are tabulated only by county, this classifica-
tion was not possible.
In a study conducted in Leadville, Colorado in
1977, Haas and Weinstein discovered that an "inter-


57
generational effect" may be important in the etiology
of low birth weight at high altitude. Comparing the
birth weights of babies born to mothers who were born
and raised at high altitude to those mothers who
migrated at a later age, they found that the native
high altitude mothers experienced a larger percentage
of low birth weight births. Although the exact loca-
tion of mother's birth was not available from the data
utilized for the present study, Table 12 corroborates
these findings, indicating that the Colorado native
mothers tend to produce a slightly higher percentage of
low birth weight babies. These Colorado results dis-
agree, however, with Haas's findings (1980) in Bolivia,
where women of high altitude Indian ancestry produced
infants with higher birth weights, regardless of
whether the mothers were born at high or low altitude.
The differences between the Colorado and Bolivian data
may indicate a genetic component in the Indian popula-
tion which confers an adaptive advantage in terms of
a mother's ability to meet the demands of pregnancy.
The short residence period at high altitude for the
Colorado population may.not have been sufficient time
for genetic adaptation to occur, thus explaining the
weight disadvantages which they experience.


CHAPTER VII
CONCLUSIONS
Colorado's infant mortality rate, once one of
the highest in the country, is now among the nation's
lowest. Previously reported increased infant mortality
rates associated with high altitude regions of the
state have now decreased to levels comparable to those
of lower altitude areas, although birth weight con-
tinues to be inversely correlated with altitude. The
lower birth weight's at high altitude are not the re-
sult of any variation in gestation time by altitude.
Comparisons of birth weight by gestational age indi-
cate that intrauterine growth retardation begins at
high altitude after week 30 of gestation. Categorizing
mortality rates by weight, low birth weight infants
have the best survival probability at high altitude.
In addition, preterm babies appear to have the best
survival probability at high altitude.
Male birth weights are consistently greater
than female birth weights, except among mortalities
occurring above 9,000'. At this level, the birth
weights of the non-surviving females are significantly
higher than the birth weights of the non-surviving >


59
males.' This information, coupled with the fact that
the female survival advantage over males is greatest
at high altitude, may indicate that females are
better adapted to the stresses of high altitude than
males.
Identifying the factors responsible for the
large state-wide reductions in both incidence of low
birth weight and infant mortality over the past 10
years, the proportionately greater decreases at high
altitude, and the increased survival of both low birth
weight and preterm infants at high altitude remain be-
yond the scope of this project. Available informa-
tion indicates that improvements in medical care and
technology, recently acquired abilities to transport
at-risk infants from isolated areas of the state to
facilities offering this improved care, and possible
variations in the relative socioeconomic composition
of .groups residing at the different altitude levels
in Colorado may be acting in concert, resulting in
these improvements in infant survival.


BIBLIOGRAPHY
Beall, Cynthia M.
1981 Optimal Birthweights in Peruvian Populations
at High and Law Altitudes. American Journal of
Physical Anthropology 56:209-216.
Blaylock, Hubert M. Jr.
1979 Social Statistics. New York: McGraw-Hill, Inc
Bloom, Bernard S.
1984 Changing Infant Mortality: The Need to Spend
More While Getting Less. Pediatrics 73:862-866.
Bowes, Watson A. Jr., George E. Fryer, Jr., and Bruce
Ellis
1984 The Use of Standardized Neonatal Mortality
Ratios to Assess the Quality of Perinatal Care
in Colorado., American Journal of Obstetrics and
Gynecology 148:1067-1073.
Cotton, Ernest K., et al.
1980 Study of Babies Born to Mothers Living at an
Altitude of 3,100 Meters. American Journal of
Obstetrics and Gynecology 138:220-222.
Frisancho, A. Roberto and Jeffrey Cossman
1970 Secular Trend in Neonatal Mortality in the
Mountain States. American Journal of Physical
Anthropology 33:103-106.
Frisancho, A. Roberto
1979 Human Adaptation A Functional Interpretation
St. Louis: C.V. Mosby.
Fraccaro, M.
1956 A Contribution to the Study of Birth Weight
Based on an Italian Sample. Annals of Human
Genetics 20:282-296.
Grahn, Douglas and Jack Kratchman
1963 Variation in Neonatal Death Rate and Birth
Weight in the United States and Possible Relations
to Environmental Radiation, Geology and Altitude.
American Journal of Human Genetics 15:329-352.


61
Guyer, Bernard, Lee Anne Wallach and Sharon L. Rosen
1982 Birth-Weight-Standardized Neonatal Mortality
Rates and the Prevention of Low Birth Weight: How
Does Massachusetts Compare with Sweden? The New
England Journal of- Medicine 306:1230-1233.
Haas, Jere D.
1980 Maternal Adaptation and Fetal Growth at High
Altitude in Bolivia. In Social and Biological
Predictors of Nutritional Status, Physical Growth
and Neurological Development. Lawrence S. Greene
and Francis E. Johnston, eds. Pp. 257-290. New
York: Academic Press.
Hock, Raymond J.
1979 The Physiology of High Altitude. In Human
Physiology and the Environment in Health and
Disease. Arthur J. Vander, ed. Pp. 103-112.
San Francisco: W.H. Freeman.
Hollingsworth, M.J.
1965 Observations on the Birth Weights and Survival
of African Babies: Single Births. Annals of
Human Genetics 28:291-300.
Karn, Mary N. and L.S. Penrose
1951 Birth Weight and Gestation Time in Relation to
Maternal Age, Parity and Infant Survival. Annals
of Eugenics 16:147-164.
Kessel, Samuel S., et al.
1984 The Changing Pattern of Low Birth Weight in
the United States. Journal of the American Medi-
cal Association 251:1978-1982.
Kitchen, W.H., et al.
1978 A Longitudinal Study of Very Low-birthweight
Infants. I: Study, Design and Mortality Rates.
Developmental Medicine and Child Neurology 20:
605-618.
Koops, Beverly L., Linda J. Morgan and Frederick C.
Battaglia
1982 Neonatal Mortality Risk in Relation to Birth
Weight and Gestational Age: Update. The Journal
of Pediatrics 101:969-977.
Lichty, John A., et al.
1957 Studies of Babies Born at High Altitude.
A.M.A. Journal of Diseases of Children 93:666-678.


62
Lubchenco, L.O., D.T. Searls and J.V. Brazie
1972 Neonatal Mortality Rate: Relationship to
Birth Weight and Gestational Age. The Journal of
Pediatrics 81:814-822.
Mazess, Richard B.
1965 Neonatal Mortality and Altitude in Peru.
American Journal of Physical Anthropology 23:
209-214.
Mazess/ Richard B.
1975 Human Adaptation to High Altitude. In Physio-
logical Anthropology. Albert Damon/ ed. Pp. 167-
209. New York: Oxford University Press.
McClung/ J.
1969 Effects of High Altitude on Human Birth.
Cambridge: Harvard University Press.
McCormick, Marie C.
1985 The Contribution of Low Birth Weight to Infant
Mortality and Childhood Morbidity. The New
England Journal of Medicine 312:82-90.
McCullough, Robert E., Joht T. Reeves and Robert L.
Liljegren
1977 Fetal Growth Retardation and Increased Infant
Mortality at High Altitude. Archives of Environ-
mental Health 32:36-39.
Minis, Jean and You Poh Seng
1954 The Effect of Age and Parity of the Mother on
Birth Weight of the Offspring. Annals of Eugenics
19:58-67.
Moore, Lorna G., et al.
1980 The Biocultural Basis of Health. St. Louis:
C.V. Mosby Company.
Moore, Lorna G., et al.
1982 Infant Birth Weight Is Related to Maternal
Arterial Oxygenation at High Altitude. Journal
Appl. Physiol.: Respirat. Environ. Exercise
Physiol. 52:695-699.
Moore, Lorna G. and Judith G. Regensteiner
1983 Adaptation to High Altitude. Annual Review
of Anthropology. 12:285-304.


63
National Center for Health. Statistics
1950-1980 Vital Statistics of the U.S. 1950-1980,
Vol. II, Mortality, Part B. Public Health Serv-
ice: Washington, D.C.
O'Donald, P.
1968 Measuring the Intensity of Natural Selection.
Nature 220:197-198.
O'Donald, P.
1970 Measuring the Change of Population Fitness by
Natural Selection. Nature 227:307-308.
Paneth, Nigel, et al.
1982 Newborn Intensive Care and Neonatal Mortality
in Law-Birth-Weight Infants. The New England
Journal of Medicine 307:149-155.
Philip, Alistair G.S., et al.
1981 Neonatal Mortality Risk for the Eighties:
The Importance of Birth Weight/Gestational Age
Groups. Pediatrics 68:122-130.
Starfield, Barbara, et al.
1982 Mortality and Morbidity in Infants with Intra-
uterine Growth Retardation. The Jounral of Ped-
iatrics 101:978-983.
Susser, Mervyn, Francis A. Marolla and Joseph Fleiss
1972 Birth Weight, Fetal Age and Perinatal Mortali-
ty. American Journal of Epidemiology 96:197-204.
Taffel, Selma
1980 Factors Associated With Low Birth Weight.
U.S. Department of Health, Education and Welfare,
National Center for Health Statistics: Hyattsville
Md.
Terrenato, L., et al.
1981 Natural Selection Associated with Birth Weight
Annals of Human Genetics 45:267-278.
Thomson, A.M. and W.Z. Billewicz
1976 The Concept of the "Light for Dates: Infant.
In The Biology of Human Fetal Growth. D.F.
Roberts and A.M. Thomson, eds. Taylor and Francis
Ltd.: London.
Ulizzi, L., M.F. Gravina and L. Terrenato
1981 Natural Selection Associated with Birth Weight
Annals of Human Genetics 45:207-212.


APPENDIX A
RAW DATA


COMPARISON OF COLORADO LOW BIRTH WEIGHT PERCENTAGES AND
INFANT MORTALITY RATES* 1969-1972 AND 1979-1982**
ALTITUDE % LOW BIRTH WEIGHT BIRTHS 1969-1972 % LOW BIRTH WEIGHT BIRTHS 1979-1982 % CHANGE 1969-1982 INFANT MORTALITY RATE 1969-1972 INFANT MORTALITY RATE 1979-1982 % CHANGE 1969-1982
9,000'+ 17.1% 13.1% 4.0% 33.6 8.6 2.5%
7-8,999' 11.3% 9.2% 2.1% 23.6 9.7 1.4%
5-6,999' 9.6% 8.1% 1.5% 18.6 9.4 .9%
3-5,999' 8.7% 6.9% 1.8% 20.7 9.1 1.2%
COLORADO 9.5% 8.0% 1.5% 19.1 9.4 1.0%
Deaths per 1,000 live births
Includes total population


RELATIONSHIPS BETWEEB ALTITUDE, BIRTH WEIGHT, SEX AND MORTALITY
COLORADO 1979-1982
ALTITUDE X BIRTH WEIGHT SURVIVING FEMALES X BIRTH WEIGHT SURVIVING MALES X BIRTH WEIGHT FEMALE INFANT MORTALITIES X BIRTH WEIGHT MALE INFANT MORTALITIES X BIRTH WEIGHT FEMALE NEONATAL MORTALITIES X BIRTH WEIGHT MALE NEONATAL MORTALITIES
9,000'+ 3001g 3142g 1975g 1588g 1786g 1496g
489 512 1022 1129 1233 1090
N=1047 N=1018 N=6 N=12 N=4 N=9
7-8,999' 3101g 3226g 1942g 221 lg 1633g 1716g
497 517 1060 841 993 917
N=4549 N=4812 N=43 N=49 N=29 N=25
5-6,999' 3177g 3312g 1999g 2036g 1593g 1586g
511 537 1114 1167 1089 1081
N=69,443 N-72,508 N=546 N=743 N=349 N=479
3-4,999' 3214g 3346g 1854g 2232g 1488g 1727g
510 552 1111 1202 1052 117
N=15,416 N=16,139 N=121 N=164 ll CO o N=111


INFANT AND NEONATAL MORTALITY RATES* BY ALTITUDE AND SEX
COLORADO 1979-1982
ALTITUDE TOTAL BIRTHS INFANT MORTALITY RATE FEMALE INFANT MORTALITY RATE MALE INFANT MORTALITY RATE NEONATAL MORTALITY RATE FEMALE NEONATAL MORTALITY RATE MALE NEONATAL MORTALITY RATE
9,000'+ 2,083 8.6 5.7 11.7 6.2 3.8 8.7
N=18 N=6 N=12 N=13 N=4 N=9
7-8,999' 9,453 9.7 9.4 10.1 5.7 6.3 5.1
N=92 N=43 N=49 N=54 N=29 N=25
5-6,999' 143,240 9.0 7.8 10.1 5.8 5.0 6.5
N=1289 N=546 N=743 N=828 N=349 N=479
3-4,999' 31,840 9.0 7.8 10.1 6.0 5.1 6.8
N=285 N=121 N=164 N=191 N=80 . N=111
COLORADO 186,616 9.0 7.9 10.1 5.8 .5.1 6.5
N=1684 N=716 N=968 N=1086 N=462 N=624
*Deaths per 1,000 live births


APPENDIX B
MAP OF COLORADO COUNTIES AND ALTITUDES


Summit = 9313'
Clear Creek = 7896'
Gilpin = 8275'
Jefferson = 5397'
Denver = 5280'
San Juan = 9318'
Lake = 10,152
MAP OF COLORADO COUNTIES AND ALTITUDES
CD
<£>