JOURNAL OF ADOLESCENT HEALTH

CARE

ORIGNAL ARTICLES

The Effect of Altitude on Adolescent Growth and

Development
ELBODORO A. FREYRE, M.D. AND MARIANELA V. ORTIZ, M.D.
 1988 EFFECT OF ALTITUDE ON ADOLESCENT GROWTH AND DEVELOPMENT
To evaluate the effect of altitude on adolescent growth
and development, three groups of healthy, well-
nourished youth of similar socioeconomic status and
ethnic gouping who resided at sea level (n = 1262
subjects), mid-altitude (n = 1743 subjects), and high
altitude (n = 1137 subjects) were studied. The
following parameters were evaluated: weight and
height in all subjects; genital and pubic hair
development in the males; and the initiation of breast
development and age of menarche in the females. At
mid-altitude, puberty appears to start at a similar age,
but lasts longer, than at sea level. At high altitude,
puberty starts significantly later and is more
prolonged than at sea level and mid-altitude. Our data
reveal relatively little effect of altitude upon growth
and weight-gain patterns or final attained heights and
weights.
KRY WORDS
Alätude and growth and development
Growth and development
Pubertal development at different altitudes
Menarche and altitude
Sexual maturation and altitude
In addition to heredity, there are other factors
capable of influencing growth and development
during childhood and adolescence (1). It has been
shown that adolescent growth and sexual
maturaNon varies in relation to race (2, 3),
nutrition (4, 5), and socioeconomic and cultural
factors (6-8), whereas it appears that climate
alone does not have any significant influence (1).
In Peru, where research on the biologic impact
of altitude has a long history (9), slow growth of
From the Centro MUico to:
de Adolescentes, Universidad Nacional de
rquests
Peni.
A. Freyre, M.D., Casilla 1328,
April 5, 1986.
144
children and adolescents as well as delayed and
prolonged pubertal development in individuals
residing at high altitude (> 3,000 meters)
compared to those living at sea level has been
reported (10—12). Children and adolescents
from the city of Puno (4,000— 5,500 meters
above sea level), when compared with subjects
living at sea level, show significantly lower
weights and heights, a slower growth curve, and
a poorly defined adolescent growth spurt (10). In
Huancayo, a city located at 3,280 meters, the
initiation of breast and penile growth, as well as
the age of menarche, are delayed, and adolescent
growth is reduced when compared to those living
at sea level (11). In the city of Cerro de Pasco
(4,200 meters), adolescents also have slower
body growth as well as delayed puberty and
menarche when compared to the residents of
Lima, at sea level (12).
The authors of these studies (10—12) explain
their findings on the basis of chronic hypoxia at
high altitude. Although attractive, this hypothesis
cannot be assumed because other concomitant
racial, nutritional, and socioeconomic variables
are also capable of influencing growth and
development (1—8).
The few studies conducted in residents from
cities of mid-alätude (1,500-3,000 meters above
sea level) showed and development to not be
significantly affected by altitude (13—15). The
growth pattern of Colombian children from
Bogotá (2,400 meters) did not reveal significant
retardation (13). In our experience in previous
studies in Arequipa, a mid-altitude city (14, 15),
male and female adoleêcents showed weights and
heights slightly below those found at sea level
(11, 12). Our males also showed a less significant
delay in their sexual maturation than that
observed in youth from higher altitudes (11, 12).
C) Society for Adolescent Medicine, 1988
 JOURNAL OF ADOLESCENT HEALTH CARE
019707008/53.50 Published by Elsevier Science Publishing Co., Inc., 52 Vanderbilt Ave., New York, NY 10017 March
 1988 EFFECT OF ALTITUDE ON ADOLESCENT GROWTH AND DEVELOPMENT
Table 1. Distribution by Age and Sex of Cross- breast development was performed only in those
Sectional Samples of Adolescents Living at Different 7—14 years of age. The physical examination
Altitudes in Perú and the estimation of the stage of sexual
maturation was performed exclusively by one of
the authors. The data from the three groups were
Group A Group B Group collected from 1978 to 1982.
C
(sea level) (2.330 m) (3,400
m)
Age Males Males Males
Females Females Females The parents were asked to complete a health
(n) (n) (n) (n) (n) (n) questionnaire containing information regarding
07 32 58 35 the length of pregnancy, birth date, birth weight,
08 46 52 42 and date of first menstruation as well as past and
09 42 54 55 59 34 48 present illnesses of their adolescent.
10 47 53 60 66 53 50 Socioeconomic status was established using a
11 68 49 105 74 56 43 point system (16) considered to be appropriate for
12 64 57 128 81 59 47 Pertl (17). The system included the following
13 n 61 124 87 62 54 indicators: per capita income, level of parents'
14 73 58 110 75 65 55 education, present occupation of the patents,
15 66 53 109 70 57 49 quality of the education center, housing
16 69 48 85 63 63 42 conditions, and family size. According to the
17 50 42 52 51 44 38 scores obtained, only individuals belonging to a
18 42 36 53 44 38 32 middle-high or a high socioeconomic class were
19 41 40 36 included in the study groups. The percentages of
20 40 42 35 individuals belonging to middle-high and high
classes were: 58% and 42% in group A; 52% and
Total 673589 963 780 602 535
48% in group B; and 60% and 40% in group C,
respectively.
Data regarding the nationality of the past two to
Materials and Methods three generafions for each family was obtained.
Three groups of children and adolescents, of both All subjects were students at "Class A" private
sexes, living in cities located at different altitudes schools or universiäes. In Peru these schools have
were studied. Group A (sea level) included 673 the highest tuitions as well as good facilities and
males and 589 females from the city of Lima; have obligatory active sports programs (17).
group B (mid-altitude), 963 males and 780 Based on the information collected from the
females from Arequipa (2,330 meters above sea questionnaires, the socioeconomic scale, and the
level); and Boup C (high altitude), 602 males and physical examination, all subjects in the three
535 females from the city of Cusco (3,400 meters groups had the following common characteristics:
above sea level) (Table 1). all were descended from at least two generations
In the three groups the ages of the males of Peruvians, all were born in Peru, all were
ranged from 9 to 20 years and females from 7 to healthy, well nourished, and born at term, all were
18 years. In all males, a complete physical permanent residents or resident for at least five
examination was performed, and the following years in their respective cities, and all belonged to
parameters were determined: weight, height a middle-high or a high socioeconomic class. The
(undressed, shoes off), and the stages of genital ethnicity of all subjects was also uniform
(G) and pubic hair (PH) development according (allöwing for the various subtle shades between
to Tanner's criteria (1). In all females, a brief white Hispanic and light Mestizo that are
physical examination with the girls paräally dominant in these socioeconomic classes in Peru.)
undressed was performed, and their weight, No Indians, blacks, Orientals, or Anglo-
height (in underwear and shoes off), and age of Caucasians were included.
menarche were determined. Because oi the For analytic purposes, the exact ages of the
reluctance of a significant number of parents to subjects were determined from the birth date to
give consent, for the evaluation of the breasts and the date of evaluation and were divided into six-
pubic hair in the older girls, a brief examination month groups. Because the number in each age
of the chest directed to establish the beginning of group was different, the means and standard

deviations for weight, height, stages of sexual
maturation, and age of menarche were estimated
from their relative frequencies (18). To estimate
the significance of the differences in mean values,
the Student t-test was used (18).
In order to evaluate possible differences
between individuals in the middle-high and high
socioeconomic class, the subjects in each group
were divided in two socioeconomic subgroups,
and the mean vai-
FRBYRBANDORTIZ
ues for weight, height, stage G2, stage B2, and
men-
arche No statistically significant
differenæs were found between the middle-high
and high subgoups.
Tabks 2 and 3 show the mean values and standard
deüäons foi weight and height of the three groups.
In the males (Table 2), at all ages, the values
for weitht and height in group A are somewhat
greater than in the öther two groups. Groups B
and C show similar values for weight at all ages
and for height up to age 11 years. Thereafter,
group C shows somewiut shorter heighb, These
differences are not statisäcally significant. Gieater
increments in height, indicaäve of a growth spurt,
were observed from 12 to 14 in groups A and B
and from 13 to 15 years in goup C.
In the females (Table 3), at all ages, the values
for hêight in groups A and B are similar and are
greater tun in goup C, but the differences are not
significant exept for age 12 years, when group C
shows lower values than the other two
groups. The values for weight are similar in the
three groups except for ages 8 and 9 years, when
group C shows sigtificantly lower values than
group A. Greater inaements in height were
JOURNAL OF ADOLESCENT HEALTH CARE
oberved from 10 to 12 years in goups A and B and
from 11 to 13 years in group c.
The weight/height rado, at all ages, for males and
vol. 9, No. 2
females in all groups, was adequate by North
American starrdards for well-nourished
individuals (19).
As shown in Table 4, in the males, the mean
ages for all stages of genital and pubic hair
development were lowest in goup A, highest in
group C, and intermediate in group B. The
differences between groups A and B for the first
Tanner G2 and PH2, were not staästielly
significant, but from G3 to PH5, the subjects in
group B attained each stage significantly later.
The youngsters in group C, as compared With
those in group B, attained all Tanner stages, from
G2 to PH5, significantly later. Because the ages
for all stages were younger in group A than in
group B, the differences between groups A and C
are also significant, including the values for G2
and PH2.
The sequence of the various Tanner stages was
identical in all groups, G2 being first and PH5 the
last. All stages of genital development were
always ahead of the respective stages of pubic
hair. The interval from the first to the last stage
(interval G2— PH5) was 4.7, 5.6, and 5.9 years
for groups A, B and C, respectively. For practical
purposes, we consider the interval G2—PH5 as
an expression of the length of the pubertal
process.
In the females, only the iniäation of breast
development (B2) and the age of menarche (M)
were detennined (Table 4). The difference
between groups A and B for the age at which B2
was attained was not statistically significant, but
it was sigtificant between groups B and C and
groups A and C. The ages for menarche were
12.2, 12.6, and 13.7 for groups A,
 1988 EFFECT OF ALTITUDE ON ADOLESCENT GROWTH AND DEVELOPMENT
Table 2. Weight and Height of Male Adolescents Living at Different Altitudes in Perú'

Group A Group 6 Group C

(sea level) (2,330 m) (3,400 m)

31.0 4.2 134.3 5.3 30.2 3.9 133.6 5.7 30.6 4.5 133.5 5.9
4.6 5.8 33.5 4.3 5.6 33.8 4.8 137.5 5.5
36.4 5.1 141.2 6.4 35.2 5.3 140.3 6.2 34.9 140.6 5.9
12 39.6 4.9 1'5.7 6.9 38.6 5.2 145.4 7.1 37.8 5.8 10.8 6.0
13 0.5 6.7 7.6 42.3 6.9 151.8 7.9 41.4 150.1 6.8
14 48.6 6.9 159.2 8.1 48.1 7.3 158.5 8.4 46.7 5.9 156.7 7.6
15 55.1 7.9 8.2 54.2 8.2 163.7 8.1 53.3 7.2 163.1 7.9
16 57.0 8.4 168.0 7.5 56.7 8.7 7.9 56.8 7.9 7.2
59.4 8.7 170.3 6.6 58.2 8.5 169.8 6.8 58.8 8.0 169.4 6.3
18 61.3 9.1 171.2 6.4 0.3 8.9 6.3 8.3 169.7 5.9
19 62.1 8.6 171.4 6.7 61.1 9.1 170.8 6.58.8 170.1 6.1
20 62.1 8.8 171.4 6.6 61.3 9.2 170.9 6.4 60.9 8.5 170.5 6.0 dffeænces in the mean values between the three goups are not
saäsäcally significant.
March

Table 3. Weight and Height of Female Adolescents Living at Different Altitudes in Perú

Group A Group B Group C

(sea level) (2,330 m) (3,400 m)

7 21.2 3.1 123.3 5.8 21.6 3.6 123.6 6.0 20.5 2.9 122.5 5.5
8 3.3 127.4 6.2 25.3 3.8 127.5 6.1 24.2" 3.5 126.8 5.8
30.2 4.4 132.0 6.4 30.0 4.9 131.8 6.6 283 4.0 130.8 6.1 10 34.0 5.36.9 34.1 4.8 137.2 6.8 33.0 4.7 135.9 6.3
11 39.1 6.27.1 37.6 5.9 143.9 7.0 37.9 5.6 141.6 6.2
12 44.7 7.0 150.6b 7.0 44.1 6.8 150.8b 7.3 43.7 6.2 6.9
13 47.6 7.5 156.1 6.8 47.4 7.4 156.4 6.9 47.9 7.1 154.4 7.0
14 49.5 7.9 157.8 49.7 8.0 157.7 6.ó 49.4 7.3 156.7 6.9
15 51.3 8.1 159.1 6.9 51.1 8.3 158.9 6.5 51.8 7.9 158.1 6.6
16 53.4 7.8 159.9 6.4 53.0 8.2 159.7 6.7 53.7 8.1 159.3 6.0
54.2 7.4 160.2 6.2 54.4 7.6 160.0 6.0 54.8 7.2 159.5 5.8 18 54.4 7.1 160.3 6.1 54.8 7.3 160.4 5.9 55.1 6.9 159.7 5.9

"Differences between goups A and C = p < 0.05.

b
Differences between groups A and C and B and C = p < 0.05.
 JOURNAL OF ADOLESCENT HEALTH CARE
B, and C, respectively, and the differences in Punc (10) consisted of Indians living in the
between groups A and B, B and C, and A and C mountains above 4,000 meters, with poor housing
were signifi„ cant. The interval from the initiation conditions, particular dietary habits, and unstated
of breast development to the age of menarche (B2 present and past health status. According to the
—M) was 1.7, 2.3, and 2.2 years for groups A, Be authors (10), the subjects showed "a slow and
and C, respectively. prolonged growth of body size and a poorly
defined adolescent stature spurt in both males and
females," but the weights and heights of their
Discussion subjects were compared with North American
Our findings suggest that at mid and high standards and with Peruvian adolescents living at
altitudes puberty is delayed in onset and/or is sea level in an urban enviroment in a study
more prolonged than at sea level in well- conducted 20 years earlier (20).
nourished, healthy adolescents from middle-high The differences in height and weight found in
and high socioeconomic classes. The differences
maintained a clear relationship with the level of
altitude.
According to our results, male adolescents
from high altitude start puberty significantly later
and have a more prolonged process of sexual
maturation than those living at sea level and at
mid-altitude. On the other hand, males who live
at mid-altitude initiate puberty at a similar age as
those at sea level, but their pubertal process is
more prolonged. The age at which the mid-
altitude youngsters attained each stage of sexual
maturation was intermediate between those at sea
level and those at high altitude.
The female data fol ¿he onset of breast
development and age of menarche also suggest
that at high altitude puberty is delayed and
prolonged and at mid-altitude is only prolonged,
when compared to girls at sea level.
Our data on body growth reveal that at sea
level male adolescents, at all ages, and the
females, for most age groups, show somewhat
greater heights and weights than those at high
altitude. The heights of the youth from mid-
alütude, for most age groups, were intermediate
between the values for sea level and high altitude.
However, the differences are discrete and not
significant, indicating that altitude has relatively
little effect on growth and weight-gain patterns or
on final attained heights and weights. In both
males and females, the subjects from high altitude
show their greatest increments in height (growth
spurt) a year later than those at sea level and mid-
altitude, which supports the theory of delayed
pubertal development.
Previous studies conducted in high-altitude
nafives and residents show comparable, but more
marked, differences in comparison to sea-level
youth (10—12). studies, howesvcr, besides
"altitude," ther»; were other concomitant factors
such as race, nutrition, and socioeconomic class
that were capable of influencing growth and
development. For example, the population studied
 1988 EFFECT OF ALTITUDE ON ADOLESCENT GROWTH AND DEVELOPMENT
148 FREYRE AND ORTIZ vol. 9, No. 2

Table 4. The Mean Ages, in Years, of the Various Stages of Sexual Maturation in Adolescents Living at Different
Alütudes in Perú

12.0 1.1 12.2 1.2 p < 0.001 13.0 0.9

12.6 0.9 12.7 1.3 p < 0.001 14.1 1.0

13.2 0.9 p < 0.001 14.1 1.1 p < 0.001 15.1 1.0

PH3 13.9 1.1 14.4 1.1 15.9 0.9

14.6 1.2 p < 0.001 15.6 1.1 p < 0.001 16.5 0.9

PH4 15.2 0.9 15.9 1.2 p < 0.001 17.2 0.8

16.2 1.1 p < 0.001 17.6 0.9 p < 0.005 18.2 1.1

PH5 16.7 1.2 p < 0.001 17.8 1.2 p < 0.001 18.9 1.2
B2b 10.5 0.9 1.1 p < 0.001 11.5 0.9

< 0.001
üreviations:
G, genital; PH, pubic hair; B, breast; M, mena:the.
for G2 and PH2 between groups A and C = p < 0.001.
'Differences for B2 between Groups A and C = p < 0.005.
these studies of individuals at sea level and at
high altitude were marked, reaching 10 kg and 15
cm in some age groups, thus suggesting the
possible presence of nutritional, environmental,
or ethnic factors. The more discrete differences
found in our study are probably due to our strict
selection of subjects. Our samples were uniform
regarding the important factors of ethnicity,
socioeconomic class, nutrition, health, housing
conditions, and oportunity for sports and social
activiäes. Lima, the capital of Peru, as well as
Arequipa and Cusco, are large ciäes where
youngsters of middle-high and high
socioeconomic classes are exposed to almost the
same enviroment and technology. The good
nutritional status of our subjects is demonstrated
by their nonnal weight-toheight ratios and by
their normal physical examinations. Also, in Peru
there is a clear relationship between good
nutrition and health in these socioeconomic
classes (16, 17).
We believe that the minor difference in the
climates of Lima, Arequipa, and Cusco is not a
p
confounding variable because climate alone has
not been shown to influence growth and
development (1).
There is a clear relationship between altitude
and ambient oxygen (21). The higher the alätude,
the lower the oxygen content of the air. The
reduction in the partial pressure of oxygen is
more significant 3,000 meters (21). This
fact defines mid-altitudes (1,500—3,000 meters
above sea level) and high altitudes (above 3,000
meters). There are several studies showing
important biologic effects of chronic hypoxia at
high altitudes (22). Although the majority of these
studies have been conducted above 3,000 meters
and relatively few have been made at lower
levels, there appears to be good evidence that the
severity of the physiologic impact depends on the
degree of hypoxia and, therefore, on the altitude
(22).
Our findings and previous reports from cities
of mid-altitude (13—15) suggest mid-alätudes
have a milder effect on growth and development
than high altitudes.
We conclude, as peviously proposed (10—12),
that altitude per se, probably because of hypoxia,

imposes a negative effect on growth and sexual
maturaäon during puberty and that the severity of
the impact has a definite relationship to the level
of altitude.
To our know: edge, there is no clear
explanation of the mechanism(s) by which
hypoxia delays growth and sexual maturation.
Llerena (12) found slight increments in the LH of
some male adoles cents from high altitude
compa:ed to those at sea level and proposed a
possible reduced testicular re sponse with a
compensatory increment in LH. Slowly
increasing testosterone levels in male youth from
high altitudes has also been described (23), but in
these two studies (12, 23), the samples were too
small to permit definite conclusions. The
possibility exists that in a state of chronic
ambient hypoxia there may be a decreased
pubertal gonadal response to gonadottopin
stimulation because in other studies abnormal
responses of the gonads (24) and adrenals (25)
have been reported in high-altitude natives and
The authors wish to thank Dr. Peter H. Body for his aid in
reviewing the manusaipt.
References
1. Tanner JM. Growth at adolescence. Oxford, Blackwell
Scientific Publishers, 2nd ed., 1962.
2. Lagka•Mierzejewgka T. Morphological and developmental
difference between negro and white Cuban youths, Hum Biol
3. Weir J, Dunn 1B, Jones EG. Race and age
Hum Biol23. Echalar E, Coyotupa J. Serum testosterone levels during pu-
8. Fleyre EA, C. Growth and development in berty in males from Lima and from Cerro de Pasco. Abstracts
of menarche. Am
J Obstet Gynecol
4. Diaz Bolio JE. The influence of nutlition on the
pubertal development in a gtmtp of female adolescents
from the city of Mérida, Yuc. a) Age of menarche. Bol
Med Hosp Infant (Méx)
JOURNAL OF ADOLESCENT HEALTH CARE
5. Diaz Bolio JE, Ramos Galván R. The influence of
nutrition on the pubertal development in a group of
female adolescents from the city of Mérida, Yuc. b)
Somatometry. Bol Med Hosp Infant (Méx) (Sp).
6. Tanner JM. Earlier maturation in man. Sci
Am 1%8;218:21— 7.
7. Susanne C. between some social and familial factors
and stature and weight of young Belgian male adults.
adolescent. Doctoral Thesis. Universidad Nacional
Mayor de San Marcos, Lima, Pert, 1971 (Sp).
12. Llerena LA. Radioimmunoassay of
luteinizing hormone. Physiologic variations and
modifications under the of altitude. Doctoral
Thesis. Universidad Peruana Cayetano Heredia, Lima,
Perú. 1973 (Sp).
13. Luna-Jaspe H, Aña J,
149
Rueda-Williamson R, et al. Cross secäcnal study of
growth, development and nutätional status in 12,138
children from Bogotá, Colombia. Arch Iaänoam Nutr
(Sp).
14.EA, Corrales de Valverde S, Rebaza R, et al. Body
growth changes in Peruvian adolescents. A comparison
with North AmeNcan standards. Trib Med (Sp).
15.EA. Normal pubertal development in males from the city
of Arequipa. High and middle socioeconomic classes.
Doctoral Thesis. Universidad Peruana Cavetano
Heredia, Lima, Perú, 1974 (Sp).
16. Meza-Cuadra A. Living conditions and health.
Lima, Perú, Editorial Causachun, 1972 (Sp).
17. Rosas-Benavidee; C, Vera M, Zeballos L. The
problem of mdnutrition in the city of Arequipa and the
role of the social worker. Bachelo Thesis. Academic
Program of Social Work, Universidad Cac61ica Santa
Marfa, Arequipap Perú, 1972 (Sp).
18. Armitage P. Statistical methods in
medical research. Oxford, Blackwell
Scientific Pub., 1971.
19. Nelson WE, ed. Textbook of Pediatrics.
Philadelphia, W.B. Saunders co., 1979.
20. Pretto JC Gómez CM. Bioanthropometric study in
Peruvian students. Boletfn del Instituto
Psicopedagógico Nacional de Lima. (Sp).
21. Comroe JH, Jr. Physiology of respiration. Chicago,
Year Book Med. Publ. Inc., 1974.
22. Porter R, Knight J, eds. A Ciba foundation
symposium on high altitude physiology: Cardiac and
respiratory aspects. Edinburgh, Churchill Livingstone,
1971.
 1988 EFFECT OF ALTITUDE ON ADOLESCENT GROWTH AND DEVELOPMENT

adolescents of different
socioeconomic condition.
Médico
9. Hurtado A.
Respiratmy adaptaäon
in the indian natives to
high altitude hypoxia.
Am J Phys Anthropol
10. Frisancho AR,
Baker VT. Altitude and
growth: A study of the
patterns of physical
growth of a high
altitude Peruvian
quechua population.
Am J Phys Anthropol
11. Peñaloza JB.
Growth and sexual
maturaäon of the
Andean
from the VI Meeting of
the Latin-American Ass.
for Research in Human
Reproduction, Lima, Perú,
1974, p. 6 (Sp).
24. Guerra-Garcfa
RA, Velasquez A,
Coyotupa J. A test of
endocrine gonadal
function in men:
Urina'.t testosterone
after the injection of
HCG. Il. A different
response of the high
altitude natives. J Clin
Endocrinol Metab
1969;29:179—82.
25. Moncloa F, Pretell
E. Corisol secretion
rate. ACTH and
opyrapon tests in high
altitude native
residents. J Clin
Endocrinol Metab