Sports Med 2006; 36 (12): 1019-1030

REVIEW ARTICLE 0112-1642/06/0012-1019/$39.95/0

© 2006 Adis Data Information BV. All rights reserved.

Adolescent Physical Activity

and Health
A Systematic Review
Pedro C. Hallal,1 Cesar G. Victora,1 Mario R. Azevedo1 and Jonathan C.K. Wells2
1 Post-Graduate Program in Epidemiology, Federal University of Pelotas, Pelotas, Brazil
2 Childhood Nutrition Research Centre, Institute of Child Health, London, UK

Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1019
1. Health Benefits of Physical Activity (PA) During Adolescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
1.1 Influence of PA During Adolescence on PA in Adulthood (Pathway A) . . . . . . . . . . . . . . . . . . . 1021
1.2 Direct Influence of PA During Adolescence on Adult Morbidity (Pathway B) . . . . . . . . . . . . . . 1021
1.3 Influence of PA in the Treatment and Prognosis of Adolescent Morbidity (Pathway C) . . . . . 1025
1.4 Influence of PA During Adolescence on Adolescent Morbidity (Pathway D) . . . . . . . . . . . . . . 1025
1.5 Other Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1026
1.6 Possible Adverse Effects of Adolescent PA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1027
2. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1027
3. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028

Abstract Physical activity in adolescence may contribute to the development of healthy

adult lifestyles, helping reduce chronic disease incidence. However, definition of
the optimal amount of physical activity in adolescence requires addressing a
number of scientific challenges. This article reviews the evidence on short- and
long-term health effects of adolescent physical activity. Systematic reviews of the
literature were undertaken using a reference period between 2000 and 2004, based
primarily on the MEDLINE/PubMed database. Relevant studies were identified
by examination of titles, abstracts and full papers, according to inclusion criteria
defined a priori. A conceptual framework is proposed to outline how adolescent
physical activity may contribute to adult health, including the following pathways:
(i) pathway A – tracking of physical activity from adolescence to adulthood; (ii)
pathway B – direct influence of adolescent physical activity on adult morbidity;
(iii) pathway C – role of physical activity in treating adolescent morbidity; and
(iv) pathway D – short-term benefits of physical activity in adolescence on health.
The literature reviews showed consistent evidence supporting pathway ‘A’,
although the magnitude of the association appears to be moderate. Thus, there is
an indirect effect on all health benefits resulting from adult physical activity.
1020 Hallal et al.

Regarding pathway ‘B’, adolescent physical activity seems to provide long-term

benefits on bone health, breast cancer and sedentary behaviours. In terms of
pathway ‘C’, water physical activities in adolescence are effective in the treatment
of asthma, and exercise is recommended in the treatment of cystic fibrosis.
Self-esteem is also positively affected by adolescent physical activity. Regarding
pathway ‘D’, adolescent physical activity provides short-term benefits; the strong-
est evidence refers to bone and mental health. Appreciation of different mecha-
nisms through which adolescent physical activity may influence adult health is
essential for drawing recommendations; however, the amount of exercise needed
for achieving different benefits may vary. Physical activity promotion must start
in early life; although the ‘how much’ remains unknown and needs further
research, the lifelong benefits of adolescent physical activity on adult health are
unequivocal.

Physical activity (PA) practice is an extremely 1. Health Benefits of Physical Activity

important, albeit difficult to measure, health-related (PA) During Adolescence
variable.[1] Individuals engaged in PA present lower
Figure 1 shows a conceptual model of how PA in
incidence rates of coronary heart disease, type 2
adolescence may be beneficial to health. The pro-
diabetes mellitus, hypertension, some cancers and
posed mechanisms include four direct effects (path-
osteoporosis.[2,3] Sedentary individuals are also
ways A–D) and three indirect effects (pathways
more likely to die prematurely than active sub-
E–G), that operate through increased PA in adults.
jects.[2] In addition to its preventive effect, PA is The evidence for the first four pathways is discussed
recommended in the treatment of several chronic in sections 1.1–1.4.
diseases.[4,5]
Although the ill effects of heart disease, osteo- Physical activity
in adolescence D
porosis and other chronic diseases manifest in adult-
hood, it is increasingly understood that their devel- C Morbidity in
A adolescence
opment starts in childhood and adolescence.[6] What B

adolescents do in their teenage years may set the

H
pattern for long periods of adulthood, as people Physical activity
in adulthood
establish many of their lifestyle choices as they G
proceed through adolescence. Therefore, promoting F Morbidity in
adulthood
PA must start early in life. However, the optimal E

amount of PA in adolescence is unclear, because I

current PA guidelines for adolescents are less clear
than those recommended for adults.[7] Mortality in
adulthood
This article is aimed at reviewing the short- and
long-term benefits of PA during adolescence for Fig. 1. The association between adolescent physical activity and
health: possible pathways. The proposed mechanisms include four
physical and mental health, based on a conceptual direct effects (pathways A–D) and three indirect effects (pathways
framework of the benefits of adolescent PA. E–G).

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
Adolescent Physical Activity and Health 1021

1.1 Influence of PA During Adolescence on on adolescent PA[8,15,18] were similar to those of the
PA in Adulthood (Pathway A) ten prospective studies.
Other methodological issues to be considered are
PA in adolescence may improve health by influ- the definitions of PA or physical fitness used in each
encing PA levels in adulthood (figure 1). The recent study and the measurement techniques. Because
literature on this issue (2000–04) was reviewed us- each indicator addresses different domains of PA,
ing the MEDLINE/PubMed database and the refer- results may vary among studies depending on the
ences and related articles of the papers identified. instrument and cutoffs used.
Combinations of the following keywords were used: The statistical approaches used in the studies
‘physical activity’, ‘physical exercise’, ‘exercise’, reviewed were variable, and this may have affected
‘physical fitness’, ‘sports practice’, ‘sports’, ‘physi- their results. The five studies[11,14,15,18,19] that treated
cal inactivity’, ‘adolescence’, ‘adolescents’, ‘teen- PA as a dichotomous variable tended to show
agers’, ‘childhood’, ‘children’, ‘adults’, ‘adult- greater effects of PA in adolescence on adult PA
hood’, ‘lifespan’, ‘maintenance’ and ‘tracking’. than the eight[8-10,12,13,16,17,20] that used continuous
Table I summarises the 13 studies[8-20] identified PA indexes. Further research is needed to explain
in terms of place, year, sample size, design, defini- this discrepancy, including investigations of the role
tions of PA in adolescence and adulthood, and main of classification errors.
results. All the 13 studies were carried out in Europe In summary, the literature suggests that PA in
or North America. adolescence is an important contributing factor to
The literature shows a consistent effect of PA adult PA levels, but existing results do not allow a
during adolescence on adult PA. However, the mag- clear recommendation on the amount of PA in ado-
nitude of this association is moderate. This result is lescence that is required to build an active lifestyle
not unexpected, since previous publications[21] in adulthood. Additional studies on this topic should
showed that adult PA is a complex behaviour, influ- address methodological issues that were raised in
enced not only by adolescent activity levels, but also the present review, including the need for studies
by socio-demographic, environmental, personal and outside Europe and North America, the advantages
behavioural variables. of prospective designs, and the need for using
standardised and consistent measurement tech-
Some methodological aspects may influence
niques, indicators and cutoffs.
study results. One might expect that studies with
long follow-up periods would report smaller effects
1.2 Direct Influence of PA During
of adolescent PA, because other factors such as
Adolescence on Adult Morbidity
work loads, injuries, chronic diseases and personal
(Pathway B)
barriers may dilute this effect over time. However,
no such trend was detected in the available literature There are few studies addressing the direct asso-
(table I). ciation between PA in adolescence and adult mor-
The timing of data collection on adolescent PA, bidity, because such investigations require long-
whether prospective or retrospective, may also af- term follow-up. In a 2001 review paper, Twisk[7]
fect study results. In the Buffalo Study,[22] highly stated that “the most classical, and perhaps the only
active adults tended to overestimate their energy study” investigating this issue was the Harvard
expenditure in adolescence. However, results from Alumni study. In 1986, Paffenbarger et al.[23] report-
the three studies that used retrospective information ed no association between adolescent PA levels and

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
© 2006 Adis Data Information BV. All rights reserved.

1022
Table I. Characteristics of the studies evaluating the influence of adolescent physical activity (PA) on adult PA

Study (Country) Sample Design Definition of PA in Definition of PA in adulthood Main results

adolescence
Hirvensalo et 1324 men and women Cohort, but participation Competitive and PA score ranging from 1 (to Competitive sports practice at 10–19y
al.[15] (Finland) aged 65–84y at in PA in adolescence recreational sports move only for minimal was associated with PA in adulthood
baseline followed up for was collected practice between necessary tasks) to 6 (exercise (OR: 1.86 in men and 2.31 in women)
8y retrospectively 10–19y to keep fitness level or doing Recreational sports practice at 10–19y
sports several days per week) was not associated with adult PA

Beunen et al.[10] 109 men followed up Cohort Physical fitness at 13, Score of sports, occupational Work and leisure-time indexes and
(Belgium) from 13 to 40y 15 and 18y and leisure-time activities, accelerometer counts in adulthood were
ranging from 3 to 15 and weakly associated with fitness in
accelerometry adolescence. The sports index in
adulthood showed a stronger
association with fitness in adolescence

Kemper et al.[16] 400 boys and girls Cohort School, work, home, School, work, home, free time, PA in adolescent presented stability
(The (baseline mean age: free time, (un)organised, sport stair coefficients of 0.29–0.35 with PA in
Netherlands) 13y) followed up for (un)organised, sport climbing and transport activities adulthood. Higher tracking values were
20y stair climbing and over the preceding 3mo observed for physical fitness
transport activities over
the preceding 3mo

Campbell et 153 boys and girls Cohort Physical work capacity Physical work capacity at a Physical work capacity showed higher
al.[12] (Canada) followed up for 12y at a heart rate of 150 heart rate of 150 bpm, daily tracking values (r = 0.24 in boys and
bpm, daily energy energy expenditure, inactive 0.46 in girls) than PA (r = 0.07–0.25 in
expenditure, inactive time and time spent in boys and 0.06–0.22 in girls)
time and time spent in moderate to vigorous PA
moderate to vigorous
PA

Kemper et al.[17] 181 boys and girls Cohort Daily PA Maximum oxygen uptake and An increase of 30% in daily PA at 13y
(The followed up from 13 to maximum slope of the track results in a 2–5% increase in aerobic
Netherlands) 27y fitness at 27y

Alfano et al.[8] 486 women (18–39y) Cross-sectional PA score ranging from Score of sports, occupational A 1-unit increase in the adolescent PA
Sports Med 2006; 36 (12)

(USA) 0 (no past sport and leisure-time activities, score was associated with an increase
participation) to 3 (high ranging from 3 to 15 of 0.22 in the adult PA score
sports participation)

Hallal et al.
Continued next page
© 2006 Adis Data Information BV. All rights reserved.

Adolescent Physical Activity and Health

Table I. Contd
Study (Country) Sample Design Definition of PA in Definition of PA in adulthood Main results
adolescence
De 172 young adults Cohort Total and moderate- Total and moderate-intensity No significant correlations were found
Bourdeaudhuij (mean age: 21y) intensity energy energy expenditure among men. Among women,
et al.[13] followed up for 7y expenditure correlations were 0.34 and 0.41 for total
(Belgium) and moderate-intensity energy
expenditure, respectively

Tammelin et 7794 men and women Cohort (data collected Sports practice after Individuals were classified as Participation in sports in adolescence
al.[19] (Finland) aged 14 and 31y by postal inquiry) school hours very active, active, moderately once a week was associated with a
active or inactive based on light greater likelihood of being active in
and brisk physical activities adulthood among women. Among men,
2–3 times per week were necessary to
have the same benefit

Kraut et al.[18] 3687 male industrial Cohort, but participation Extracurricular Leisure-time PA. Subjects were Participation in sports in adolescence
(Israel) workers in PA in adolescence organised sports classified as active if increased the likelihood of being active
was collected activities performing at least 1 day per in adulthood (OR: 3.6)
retrospectively week of PA during at least 30
min each

Trudeau et 166 subjects aged Cohort Total weekly time spent Total weekly time spent on PA Childhood PA was positively associated
al.[20] (Canada) 10–12y at baseline and on PA with total adult PA (r = 0.20)
followed up for ~24y

Beunen et al.[9] 166 men followed from Cohort Weekly time spent in Activity counts using A regression model including weekly
(Belgium) 1969 to 1996 sports activities accelerometer time spent in sports activities, flexibility,
pulse recovery, and degree or
urbanisation at 16y explained 12.9% of
activity counts at 40y

Gordon-Larsen 13 030 men and Cohort Individuals were Individuals were classified as Two-thirds of the adolescents who
et al.[14] (USA) women followed from classified as achieving achieving or not achieving ≥5 achieved ≥5 sessions of moderate to
1994 to 2002 or not achieving ≥5 weekly sessions of moderate or vigorous PA remained achieving this
weekly sessions of vigorous PA threshold in adulthood
moderate to vigorous
PA
Sports Med 2006; 36 (12)

Boreham et 476 men and women Cohort Daily participation in Score of sports, occupational PA in adolescence was positively
al.[11] (Northern visited in adolescence activities that were and leisure-time activities, associated with PA in adulthood in
Ireland) (mean age: 15y) and based around a typical ranging from 3 to 15 males (weighted κ: 0.20). In females,
early adulthood (mean school day no such trend was observed (weighted
age: 22y) κ: 0.02)

1023
bpm = beats/min; OR = odds ratio.
1024 Hallal et al.

the incidence of cardiovascular disease in adult- some evidence that exercise-induced gains in bone
hood. These results led to the notion that PA gains mass in children are maintained into adulthood,
are lost easily, and formerly active subjects would suggesting that PA habits during childhood may
present similar morbidity risks as those who were have long-lasting benefits on bone health.”
consistently sedentary. Other studies, however, ap- In terms of risk factors for cardiovascular diseas-
peared more recently with more positive evidence of es, three prospective studies[28-30] found no associa-
long-term benefits of adolescent PA on adult health. tions between PA in adolescence and several out-
A literature search for papers on this issue comes, including body fat, fat distribution, blood
(2000–04) was carried out using the keywords pressure, cardiorespiratory fitness, total cholesterol,
‘physical activity’, ‘physical exercise’, ‘exercise’, high-density lipoprotein-cholesterol (HDL-C) and
‘physical fitness’, ‘sports practice’, ‘sports’, ‘physi- triglycerides. Recently, Hernelahti et al.[31] reported
cal inactivity’ combined with ‘adolescence’, ‘ado- that aerobic exercise practice in adolescence re-
lescents’, ‘teenagers’, ‘childhood’, ‘children’, duced diastolic blood pressure in adulthood. There-
‘adults’ and ‘adulthood’. These keywords were also fore, most of the literature does not report a positive
combined with several outcomes: ‘cancer’, ‘osteo- long-term impact of PA on risk factors for cardio-
porosis’, ‘fractures’, ‘coronary heart disease’, ‘car- vascular morbidity. However, this apparent lack of
diovascular disease’, ‘blood pressure’, ‘high blood effect may be due to measurement problems. For
pressure’, ‘glucose levels’, ‘diabetes’, ‘body mass example, the same Irish study found no effect of PA,
index’, ‘body fatness’, ‘overweight’, ‘obesity’, ‘de- but reported a significant association between physi-
pression’, ‘anxiety’, ‘self-esteem’ and ‘mental cal fitness in adolescence and beneficial changes in
health’. Nine relevant studies were identified, in- serum total cholesterol and HDL-C concentrations,
cluding four literature reviews. systolic and diastolic blood pressure and body fat-
In a comprehensive review on early exposures ness in adulthood[28]
for breast cancer, Okasha et al.[24] concluded that 16 Some researchers studied sedentary behaviours
case-control studies suggest that PA in adolescence rather than PA. In a prospective birth cohort study,
decreases the risk of breast cancer, although data Hancox et al.[32] found that child and adolescent
from three cohort studies do not support this finding. television viewing was associated with higher body
The authors conclude that an effect is likely, but mass index, lower cardiorespiratory fitness, in-
were unable to quantify the amount of PA required creased cigarette smoking and raised serum choles-
to provide such benefit, because the studies re- terol in adulthood.
viewed used inconsistent definitions, cutoffs and In summary, the literature indicates a consistent
instruments for measuring PA. long-term protective effect of adolescent PA on
In a recent review, Karlsson[25] concluded that bone health. Results for breast cancer are also most-
adolescent PA may reduce fracture risk in later life, ly positive. In terms of risk factors for cardiovascu-
even if activity levels are reduced in adulthood. lar diseases, most findings are negative. However,
However, the author points out that the quality of the sedentary behaviour during childhood and adoles-
evidence is still weak. In a 2000 review, Khan et cence, as well as poor physical fitness in adoles-
al.[26] concluded that adolescent PA plays a vital role cence, were both found to be associated with poor
in optimising peak bone mass, and that its benefits adult health outcomes, each in a single study.
may extend into adulthood. The American College There is a clear need for additional cohort stud-
of Sports Medicine[27] recently stated that “there is ies. Randomised trials would be ideal, but they

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
Adolescent Physical Activity and Health 1025

would require long-term follow-up and would have young adult females showed that exercise was more
to deal with the ethical issue of offering extra PA to effective than cognitive-behavioural therapy in re-
some adolescents but not to others. Again, limited ducing pursuit of thinness, change in body composi-
data are available from low- and middle-income tion, frequency of bingeing, purging and laxative
countries. The available evidence does not contrib- abuse.[37] Further studies on this issue are warranted,
ute concretely to the establishment of PA guidelines in order to confirm these findings.
for adolescents, based on the long-term effects on A systematic review,[38] including six studies,
adult health. found that PA significantly improved forced vital
capacity among cystic fibrosis patients. Changes in
1.3 Influence of PA in the Treatment and other lung function parameters showed similar
Prognosis of Adolescent Morbidity trends, but these were not statistically significant. In
(Pathway C) summary, the literature shows consistent benefits of
swimming (but not of PA per se) in the treatment of
The optimal design for assessing the effect of PA asthma. Adolescent PA also seems to improve self-
on the prognosis and treatment of specific adoles- esteem, and to increase lung function among cystic
cent diseases is the randomised controlled trial. The fibrosis patients. Further research is needed to clari-
MEDLINE/PubMed (2000–04) and Cochrane fy the role of PA in the treatment of obesity and
databases were searched for randomised controlled bulimia. Many of the randomised trials on this issue
trials and reviews, respectively, including adoles- are affected by methodological drawbacks, particu-
cents (aged <20 years), and using the keywords: larly small sample sizes. Although this pathway is as
‘physical activity, ‘physical exercise’, ‘exercise’, important as the others described in this review,
‘physical fitness’, ‘sports practice’, ‘sports’ and recommendations of adolescent PA should not be
‘physical inactivity’. based on unhealthy subjects.
A systematic review of eight randomised con-
trolled trials including individuals with asthma aged 1.4 Influence of PA During Adolescence on
≥8 years concluded that physical training does not Adolescent Morbidity (Pathway D)
alter resting lung function, nor the number of days
per week with wheezing.[33] However, another sys- Adolescent PA may have a direct influence on
tematic review concluded that swimming practice morbidity risk in adolescence (figure 1). In a 2001
reduces the severity of asthma symptoms.[34] review of the literature, Twisk[7] divided the short-
The role of PA on the treatment of adolescent term benefits of adolescent PA on adolescent mor-
obesity was also evaluated in a systematic review[35] bidity into three main outcomes: cardiovascular dis-
including 18 randomised controlled trials. The au- ease, bone health and emotional conditions. The
thors were unable to reach a firm conclusion about author concluded that adolescent PA: (i) had no
this topic, largely due to methodological shortcom- consistent effect on lipid levels, blood pressure or
ings of the original studies. glucose levels in adolescence; (ii) was positively
A systematic review,[36] including 23 studies, related with HDL-C and cardiorespiratory fitness,
found that exercise in childhood and adolescence and negatively with body fatness; (iii) improved
has positive short-term effects on self-esteem, al- peak bone mass; and (iv) raised self-esteem and
though most studies are small and of poor quality. lowered stress levels in adolescents.
A randomised trial on the effect of exercise on To update the Twisk review, the MEDLINE/
the treatment of bulimia among adolescent and PubMed database was researched for papers pub-

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
1026 Hallal et al.

lished between 2000 and 2004 not included in the of conditions such as obesity and asthma. Therefore,
previous review.[7] The following keywords were a cross-sectional association between PA and such
used: ‘physical activity’, ‘physical exercise’, ‘exer- conditions may only reflect compliance with profes-
cise’, ‘physical fitness’, ‘sports practice’, ‘sports’ or sional advice. Prospective studies, including adoles-
‘physical inactivity’, in combination with either cents who at baseline do not have the condition of
‘peak bone mass’, ‘coronary heart disease’, ‘cardio- interest, are required to sort out these effects.
vascular disease’, ‘blood pressure’, ‘high blood
Because chronic disease morbidity in adoles-
pressure’, ‘glucose levels’, ‘diabetes’, ‘body mass
cence is uncommon, some authors have studied the
index’, ‘body fatness’, ‘overweight’, ‘obesity’, ‘de-
effect of PA in adolescence on risk factors. Howev-
pression’, ‘anxiety’, ‘self-esteem’ or ‘mental
er, PA in adolescence may reduce adult morbidity
health’. Seven research papers and one review were
even though it does not appear to affect risk factors
identified.
measurable during adolescence itself. This is why in
A prospective study including 4594 adolescents our conceptual model, we propose two separate
concluded that PA was inversely related to depres- mechanisms: pathway D represents effects that are
sive symptoms in early adolescence.[39] Another trial mediated through risk factors or diseases that are
showed that fourth grade students submitted to a detected in adolescence, whereas pathway B in-
6-week aerobic exercise programme were less likely cludes a direct effect on adult morbidity (or at least
to present depression and presented better self-es- an effect for which the mediating factors are not
teem than controls, but exercise did not reduce anxi- measurable during adolescence).
ety.[40] Another randomised trial showed that PA
reduces anxiety sensitivity, a precursor of panic There are some important research gaps in this
attacks and panic disorder.[41] literature. Again, data from low- and middle-income
countries are rare. There is a need for additional
A large-scale randomised trial, including 1140
studies, particularly on topics such as cardiovascular
adolescents (aged 11–14 years), showed that exer-
health for which there is no consistency in the litera-
cise lowered systolic and diastolic blood pressure.
ture. Definition of PA guidelines for adolescents
Fat mass was reduced, but body mass index was not
based on the short-term effects on health is not
affected.[42]
possible at this stage. The amount of PA required for
Regarding bone health, Yannakoulia et al.[43] providing short-term health benefits seems to vary
found dance to be a protective factor against low according to the disease or condition.
bone density. In a prospective study, sports practice
was associated with a 7-fold protection against low
bone density in school-aged girls.[44] A school-based 1.5 Other Pathways
jumping intervention improved bone mineral con-
tent in pubertal girls.[45] The American College of Figure 1 also shows other pathways (E–G) link-
Sports Medicine[27] stated in 2004 that PA in adoles- ing PA in adults to morbidity, prognosis and mortal-
cence exerts both short- and long-term benefits on ity, or that reflect progression from adolescent mor-
bone health. bidity to adult morbidity (pathway H) or from the
Most studies reviewed above used a cross-sec- latter to mortality (pathway I). Reviewing the evi-
tional design. This may lead to reverse causality. As dence for these pathways is beyond the scope of the
discussed in section 1.3, PA is used in the treatment present review.

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
Adolescent Physical Activity and Health 1027

1.6 Possible Adverse Effects of not usually considered in drawing recommenda-

Adolescent PA tions.
This discussion is even more relevant to adoles-
PA during adolescence can also cause harm. Be- cents, for whom the short-term effects of PA may be
ing forced to exercise in childhood and adolescence evident, but long-term effects will only become
has been associated with inactivity in adulthood.[46] apparent after several decades. Twisk[7] proposed
A comprehensive review showed that early sports that PA guidelines for children and adolescents must
specialisation, as in the case of competitive athletes, not be solely based on physical health benefits. We
is associated with higher dropout rates, worsened expand this view by showing that there are several
performance and motor development in adulthood, pathways by which adolescent PA may be beneficial
as well as with a higher risk of injuries.[47] Eating for physical and mental health. Although a number
disorders are receiving increasing attention particu- of relevant studies are emerging, definition of ado-
larly among female adolescents, and such conditions lescent PA threshold values is still a challenge.
often include exaggerated PA.[48] Strenuous PA may Previous adolescent PA guidelines[52] recom-
affect the female reproductive system and lead to mended that, in addition to daily activities, youths
‘athletic amenorrhoea’.[49] Weight lifting during pu- should practice at least 20 minutes per day of moder-
berty can cause serious musculoskeletal injuries, ate to vigorous PA, three times or more each week.
such as ruptured intervertebral discs, spondylolysis In 1998, new recommendations advised daily prac-
and spondylolisthesis, fractures and meniscal inju- tice of moderate to vigorous PA for approximately
ries of the knee, as well as interrupting growth.[50] 60 minutes.[53] Additional activities, at least twice a
week, were recommended as desirable for improv-
2. Discussion ing bone health, resistance and strength.[53] A recent
comprehensive review[54] confirmed these guide-
The present review has shown that, whereas PA lines are appropriate, after examining 850 papers.
in adolescence does not seem to have all the benefi- Although most interventions used 30–45 minutes
cial effects that have been claimed in recent years, it per section, 3–5 days per week, the authors conclud-
has unequivocal advantages that justify its strong ed that “school-age youth should participate daily in
promotion from the public health viewpoint. The 60 minutes or more of moderate to vigorous physi-
broad scope of potential advantages arising from cal activity that is developmentally appropriate, en-
adolescent PA, highlighted in this review, require a joyable, and involves a variety of activities.”
rethinking of current recommendations. A crucial point is that the thresholds for the
Cutoffs for adults[3,51] are based on PA thresholds benefits of adolescent PA may vary considerably
beyond which there is a measurable reduction in depending on the health condition under study. The
morbidity and mortality. In these recommendations, amount of PA required for improving self-esteem,
thresholds were largely based on the incidence and for example, is not necessarily the same required to
mortality of coronary heart disease, assessed in pro- decrease body fat. However, separate recommenda-
spective cohort studies. These outcomes typically tions for each outcome would be excessively de-
have long induction periods, and thus the effects of tailed and may lead to confusion. It is also important
PA on health may take years or decades to become to highlight that guidelines for adolescent PA should
evident. On the other hand, there are also short-term be based on healthy subjects, and therefore, evi-
effects of PA, for example on mental health, that are dence from pathway ‘C’ should not be used for

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
1028 Hallal et al.

drawing recommendations. Finally, although studies may have harmful effects on growing adolescents,
on physical fitness are of interest for understanding the possibility of recommending an upper limit for
the relationships between fitness and health, guide- the amount of PA should be discussed. Promotion of
lines should focus on PA rather than fitness. PA must start as early as possible, and needs to be
Definition of adolescent PA guidelines is beyond kept as a public health priority. Although the ‘how
the scope of this article, but our conceptual frame- much’ remains unknown and deserves further re-
work, and the recognition that domains of PA are search, the benefits of adolescent PA on adult health
different from those of adults may help governmen- are unequivocal.
tal and non-governmental agencies involved in cre-
ating these guidelines. Acknowledgements

No sources of funding were used to assist in the prepara-

3. Conclusions tion of this review. The authors have no conflicts of interest
that are directly relevant to the content of this review.
This article provided a framework for under-
standing the potential benefits from adolescent PA,
and reviewed the evidence for each of the links in References
1. LaPorte RE, Montoye HJ, Caspersen CJ. Assessment of physi-
this framework. We found that (see figure 1): cal activity in epidemiologic research: problems and prospects.
Public Health Rep 1985; 100: 131-46
• There is consistent evidence that adolescent PA 2. Bauman AE. Updating the evidence that physical activity is
is positively associated with adult PA levels. good for health: an epidemiological review 2000-2003. J Sci
• Adolescent PA provides a long-term protective Med Sport 2004; 7: 6-19
3. US Department of Health and Human Services. Physical activi-
effect on bone health. Most evidence on the role ty and health: a report from the Surgeon General. Atlanta
of adolescent PA on breast cancer is positive. (GA): National Center for Chronic Disease Prevention and
Sedentary behaviour during childhood and ado- Health Promotion, 1996
4. American Diabetes Association. Clinical practice recommenda-
lescence, as well as poor physical fitness in ado- tions. Diabetes Care 2001; 24: S33-43
lescence, were both found to be associated with 5. Joint National Committee on Prevention, Detection and Treat-
ment of High Blood Pressure. The Sixth Report of the Joint
poor adult health outcomes. National Committee on Prevention, Detection and Treatment
• There are consistent benefits of swimming (but of High Blood Pressure. Bethesda (MD): National Institutes of
not of PA per se) in the treatment of asthma. Health, National Heart Lung and Blood Institute, 1997
6. Parsons TJ, Power C, Logan S, et al. Childhood predictors of
Adolescent PA also seems to improve self-es- adult obesity: a systematic review. Int J Obes Relat Metab
teem, and to increase lung function among cystic Disord 1999; 23 Suppl. 8: S1-107
7. Twisk JW. Physical activity guidelines for children and adoles-
fibrosis patients. cents: a critical review. Sports Med 2001; 31: 617-27
• Adolescent PA provides short-term benefits; the 8. Alfano CM, Klesges RC, Murray DM, et al. History of sport
strongest evidence refers to bone and mental participation in relation to obesity and related health behaviors
in women. Prev Med 2002; 34: 82-9
health. 9. Beunen GP, Lefevre J, Philippaerts RM, et al. Adolescent
Definition of evidence-based PA guidelines for correlates of adult physical activity: a 26-year follow-up. Med
Sci Sports Exerc 2004; 36: 1930-6
adolescents is still a public health challenge, but 10. Beunen GP, Philippaerts RM, Delvaux K, et al. Adolescent
appreciation of all the mechanisms by which adoles- physical performance and adult physical activity in Flemish
males. Am J Hum Biol 2001; 13: 173-9
cent PA may influence adult health is warranted.
11. Boreham C, Robson PJ, Gallagher AM, et al. Tracking of
Guidelines should take into account the fact that physical activity, fitness, body composition and diet from
there are multiple benefits of adolescent PA and that adolescence to young adulthood: the Young Hearts Project,
Northern Ireland. Int J Behav Nutr Phys Act 2004; 1: 14
the amount of exercise needed for reaching different 12. Campbell PT, Katzmarzyk PT, Malina RM, et al. Prediction of
benefits may vary. Finally, because excessive PA physical activity and physical work capacity (PWC150) in

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
Adolescent Physical Activity and Health 1029

young adulthood from childhood and adolescence with consid- Sports Study – an eight-year follow-up study. Int J Sports Med
eration of parental measures. Am J Hum Biol 2001; 13: 190-6 2002; 23 Suppl. 1: S27-31
13. De Bourdeaudhuij I, Sallis J, Vandelanotte C. Tracking and 30. Lefevre J, Philippaerts R, Delvaux K, et al. Relation between
explanation of physical activity in young adults over a 7-year cardiovascular risk factors at adult age, and physical activity
period. Res Q Exerc Sport 2002; 73: 376-85 during youth and adulthood: the Leuven Longitudinal Study on
14. Gordon-Larsen P, Nelson MC, Popkin BM. Longitudinal physi- Lifestyle, Fitness and Health. Int J Sports Med 2002; 23 Suppl.
cal activity and sedentary behavior trends: adolescence to 1: S32-8
adulthood. Am J Prev Med 2004; 27: 277-83 31. Hernelahti M, Levalahti E, Simonen RL, et al. Relative roles of
15. Hirvensalo M, Lintunen T, Rantanen T. The continuity of physi- heredity and physical activity in adolescence and adulthood on
cal activity: a retrospective and prospective study among older blood pressure. J Appl Physiol 2004; 97: 1046-52
people. Scand J Med Sci Sports 2000; 10: 37-41 32. Hancox RJ, Milne BJ, Poulton R. Association between child and
16. Kemper HC, de Vente W, van Mechelen W, et al. Adolescent adolescent television viewing and adult health: a longitudinal
motor skill and performance: is physical activity in adoles- birth cohort study. Lancet 2004; 364: 257-62
cence related to adult physical fitness? Am J Hum Biol 2001;
33. Ram FS, Robinson SM, Black PN. Physical training for asthma.
13: 180-9
Cochrane Database Syst Rev 2000: CD001116
17. Kemper HC, Twisk JW, Koppes LL, et al. A 15-year physical
34. Rosimini C. Benefits of swim training for children and adoles-
activity pattern is positively related to aerobic fitness in young
cents with asthma. J Am Acad Nurse Pract 2003; 15: 247-52
males and females (13-27 years). Eur J Appl Physiol 2001; 84:
395-402 35. Summerbell CD, Ashton V, Campbell KJ, et al. Interventions
18. Kraut A, Melamed S, Gofer D, et al. Effect of school age sports for treating obesity in children. Cochrane Database Syst Rev
on leisure time physical activity in adults: the CORDIS Study. 2003: CD001872
Med Sci Sports Exerc 2003; 35: 2038-42 36. Ekeland E, Heian F, Hagen KB, et al. Exercise to improve self-
19. Tammelin T, Nayha S, Hills AP, et al. Adolescent participation esteem in children and young people. Cochrane Database Syst
in sports and adult physical activity. Am J Prev Med 2003; 24: Rev 2004: CD003683
22-8 37. Sundgot-Borgen J, Rosenvinge JH, Bahr R, et al. The effect of
20. Trudeau F, Laurencelle L, Shephard RJ. Tracking of physical exercise, cognitive therapy, and nutritional counseling in treat-
activity from childhood to adulthood. Med Sci Sports Exerc ing bulimia nervosa. Med Sci Sports Exerc 2002; 34: 190-5
2004; 36: 1937-43 38. Bradley J, Moran F. Physical training for cystic fibrosis. Coch-
21. Trost SG, Owen N, Bauman AE, et al. Correlates of adults’ rane Database Syst Rev 2002:CD002768
participation in physical activity: review and update. Med Sci 39. Motl RW, Birnbaum AS, Kubik MY, et al. Naturally occurring
Sports Exerc 2002; 34: 1996-2001 changes in physical activity are inversely related to depressive
22. Falkner KL, McCann SE, Trevisan M. Participant characteris- symptoms during early adolescence. Psychosom Med 2004;
tics and quality of recall of physical activity in the distant past. 66: 336-42
Am J Epidemiol 2001; 154: 865-72 40. Crews DJ, Lochbaum MR, Landers DM. Aerobic physical ac-
23. Paffenbarger RS, Hyde RT, Wing AL, et al. Physical activity, tivity effects on psychological well-being in low-income His-
all-cause mortality, and longevity of college alumni. N Engl J panic children. Percept Mot Skills 2004; 98: 319-24
Med 1986; 314: 605-13 41. Broman-Fulks JJ, Berman ME, Rabian BA, et al. Effects of
24. Okasha M, McCarron P, Gunnell D, et al. Exposures in child- aerobic exercise on anxiety sensitivity. Behav Res Ther 2004;
hood, adolescence and early adulthood and breast cancer risk: 42: 125-36
a systematic review of the literature. Breast Cancer Res Treat
42. McMurray RG, Harrell JS, Bangdiwala SI, et al. A school-based
2003; 78: 223-76
intervention can reduce body fat and blood pressure in young
25. Karlsson MK. Physical activity, skeletal health and fractures in adolescents. J Adolesc Health 2002; 31: 125-32
a long term perspective. J Musculoskelet Neuronal Interact
43. Yannakoulia M, Keramopoulos A, Matalas AL. Bone mineral
2004; 4: 12-21
density in young active females: the case of dancers. Int J Sport
26. Khan K, McKay HA, Haapasalo H, et al. Does childhood and
Nutr Exerc Metab 2004; 14: 285-97
adolescence provide a unique opportunity for exercise to
strengthen the skeleton? J Sci Med Sport 2000; 3: 150-64 44. Ford MA, Bass MA, Turner LW, et al. Past and recent physical
27. Kohrt WM, Bloomfield SA, Little KD, et al. American College activity and bone mineral density in college-aged women. J
of Sports Medicine Position Stand: physical activity and bone Strength Cond Res 2004; 18: 405-9
health. Med Sci Sports Exerc 2004; 36: 1985-96 45. MacKelvie KJ, Khan KM, Petit MA, et al. A school-based
28. Boreham C, Twisk J, Neville C, et al. Associations between exercise intervention elicits substantial bone health benefits: a
physical fitness and activity patterns during adolescence and 2-year randomized controlled trial in girls. Pediatrics 2003;
cardiovascular risk factors in young adulthood: the Northern 112: e447
Ireland Young Hearts Project. Int J Sports Med 2002; 23 46. Taylor WC, Blair SN, Cummings SS, et al. Childhood and
Suppl. 1: S22-6 adolescent physical activity patterns and adult physical activi-
29. Hasselstrom H, Hansen SE, Froberg K, et al. Physical fitness ty. Med Sci Sports Exerc 1999; 31: 118-23
and physical activity during adolescence as predictors of car- 47. Stricker PR. Sports training issues for the pediatric athlete.
diovascular disease risk in young adulthood: Danish Youth and Pediatr Clin North Am 2002; 49: 793-802

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)
1030 Hallal et al.

48. Beumont PJ, Arthur B, Russell JD, et al. Excessive physical 53. Biddle S, Cavill N, Sallis J. Young and Active? Young people
activity in dieting disorder patients: proposals for a supervised and health-enhancing physical activity – evidence and implica-
exercise program. Int J Eat Disord 1994; 15: 21-36 tions. London: Health Education Authority, 1998
49. Eliakim A, Beyth Y. Exercise training, menstrual irregularities 54. Strong WB, Malina RM, Blinkie CJ, et al. Evidence based
and bone development in children and adolescents. J Pediatr physical activity for school age youth. J Pediatr 2005; 146:
Adolesc Gynecol 2003; 16: 201-6 732-7
50. Risser WL. Weight-training injuries in children and adolescents.
Am Fam Physician 1991; 44: 2104-8
51. Pate RR, Pratt M, Blair SN, et al. Physical activity and public Correspondence and offprints: Pedro C. Hallal, Post-Gradu-
health: a recommendation from the Centers for Disease Con-
ate Program in Epidemiology, Federal University of Pelo-
trol and Prevention and the American College of Sports
Medicine. JAMA 1995; 273: 402-7 tas, Av. Duque de Caxias 250, Pelotas RS, CEP: 96030-002,
52. Sallis J, Patrick K. Physical activity guidelines for adolescents: Brazil.
consensus statement. Ped Exerc Sci 1994; 6: 302-14 E-mail: [email protected]

© 2006 Adis Data Information BV. All rights reserved. Sports Med 2006; 36 (12)