See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/343892279

Factors associated with physical activity levels in late adolescence: A

prospective study

Article in Medycyna Pracy · August 2020

DOI: 10.13075/mp.5893.01012

CITATIONS READS

8 744

3 authors, including:

Damir Sekulic Tine Sattler

University of Split University of Ljubljana
212 PUBLICATIONS 4,923 CITATIONS 24 PUBLICATIONS 1,278 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Damir Sekulic on 28 August 2020.

The user has requested enhancement of the downloaded file.

Medycyna Pracy 2020;71(6)
http://medpr.imp.lodz.pl/en https://doi.org/10.13075/mp.5893.01012
ORIGINAL PAPER

FACTORS ASSOCIATED WITH PHYSICAL ACTIVITY LEVELS

IN LATE ADOLESCENCE: A PROSPECTIVE STUDY
Damir Sekulic1, Jelena Rodek1, Tine Sattler2
1
University of Split, Split, Croatia
Faculty of Kinesiology
2
University of Ljubljana, Ljubljana, Slovenia
Faculty of Sport

Abstract
Background: Reaching an appropriate physical activity level (PAL) in adolescence is an important public health problem. This
study aimed to evaluate factors associated with PAL and changes in PAL in late adolescence. Material and Methods: The sample
involved 411 adolescents (201 girls) who were 16 years of age at study baseline; the participants completed a structured validated
questionnaire at baseline (the beginning of the third grade of high school) and again at follow-up (the end of the fourth grade; when
they were 18 years of age). Variables were obtained at both testing waves and included the following predictors: sociodemographic
indices (sex, socioeconomic status, parental education), sports factors (participation in individual and team sports, sports achieve-
ment, experience in sports), and PAL (the primary outcome), which was measured using the Physical Activity Questionnaire for
Adolescents. Results: A significant decrease in PAL between baseline and follow-up testing was evidenced (t-test = 6.17, p < 0.001).
A logistic regression model calculated with a dichotomized outcome (normal PAL vs. low PAL), and sex as a covariate, showed
a significant influence of participation in team sports and maternal education on PAL, both at baseline (OR: 1.56, 95% CI: 1.11–1.87;
OR: 1.38, 95% CI: 1.02–1.90) and at follow-up (OR: 1.45, 95% CI: 1.01–1.90; OR: 1.35, 95% CI: 1.08–1.70, for team sports and ma-
ternal education, respectively). Conclusions: The study confirmed certain associations between the studied variables and PAL, but
there was no significant influence of the observed indicators on changes in PAL in late adolescence. Further studies evaluating other
predictors of changes in PAL are warranted. Med Pr. 2020;71(6)
Key words: predictors, sport, physical activity, puberty, sociodemographics, physical literacy

Corresponding author: Damir Sekulic, University of Split, Faculty of Kinesiology, Teslina 6, 21000 Split, Croatia,
e-mail: [email protected]
Received: May 16, 2020, accepted: June 15, 2020

INTRODUCTION Recently, a prospective study demonstrated chang-

es and factors of influence in PAL among adolescents
Physical activity (PA) directly and indirectly influences from Bosnia and Herzegovina that occurred over a pe-
health-related indices and is considered an important riod of 2 years [6]. Predictors were sociodemograph-
determinant of the overall health status [1]. Apart from ic characteristics, and variables of substance misuse
being related to the development of obesity, a low physi- (consumption of cigarettes, alcohol and illicit drugs),
cal activity level (PAL) is known to be connected to var- and PAL (outcome) was measured using the Physical
ious non-communicable diseases, such as hypertension, Activity Questionnaire for Adolescents (PAQ-A). Apart
cardiovascular diseases, and some types of cancer [2]. from several correlations between the studied predic-
Although most of the negative health effects connect- tors and changes in PAL that occurred during the study
ed with low PAL develop later in life, lifelong patterns period (at 16–18 years of age), the authors concluded
related to health behavior are established during child- that certain relationships should be contextualized in
hood and adolescence [3]. Therefore, the promotion of light of the fact that in the studied period a large num-
reaching an appropriate PAL should begin at an early ber of children quit competitive sports, which simulta-
age [4]. For that reason, there is growing interest in re- neously influenced changes in the sociocultural envi-
search focusing on factors associated with PAL in differ- ronment, and PAL. However, it is important to note that
ent periods of life, and one of the promising approach- this study did not specifically observe (measure) sports
es is the identification of the factors associated with factors, and the stated interpretations were based solely
“changes in PAL” over a certain period of time [5,6]. on the authors’ own experience.
2 D. Sekulic et al. Nr 6

While PA includes any bodily movement produced MATERIAL AND METHODS

by skeletal muscles, physical exercise (which is most-
ly nowadays evidenced by participation in sports ac- Participants and design of the study
tivities) refers to planned and structured PA per- The sample of participants was selected in the terri-
formed with some specific reason (i.e., energy ex- tory of southern Croatia (specifically Split–Dalmatia
penditure, improvement of body function, competi- County). The total sample included 411 adolescents
tive result) [7]. Logically, sports participation is con- (201 girls) who were 16.2±1.3 years old and attending
sidered as important factor which directly contrib- their third year of high school at study baseline (the
utes to PAL, particularly among children and adoles- first wave of testing). The second wave of testing (fol-
cents [8]. The participation in organized (competi- low-up measurement) commenced at the end of their
tive) sports is “physically demanding,” and the over- fourth year of high school, when the participants were
all PAL of those children involved in sports is regu- approximately 18 years of age. The sampling procedure
larly found to be higher than that of non-athletic chil- included several phases. First, the authors stratified all
dren [9–11]. As a result, it is not surprising that stud- high schools in the territory of Split–Dalmatia County
ies have reported a positive correlation between par- into 2 clusters according to the size of the schools.
ticipation in sports and PAL, emphasizing the impor- Then, one-third of the high schools clustered in each
tance of participation in organized sports activities in group were randomly selected. By the next phase, one-
reaching an appropriate PAL in youth [9]. This prob- half of the third grade classes in each school were ran-
lem is additionally important because of the global- domly selected. For those schools where there was
ly confirmed significant decrease in PAL in adoles- only one third grade class, the random selection was
cence [12]. Therefore, studies identifying factors as- not done and the one existing class was included in
sociated with PAL changes in adolescence are essen- the study. After contacting the school authorities and
tial, because any information that will help in reduc- obtaining permission for the study, parental informed
ing the negative trends in PA changes will be highly consent was obtained. At the first wave of testing,
beneficial, both from a scientific perspective and from 441 participants were tested, but only those who par-
a public health perspective. ticipated in both testing waves were included in this
With regard to the association between sports par- investigation.
ticipation and PAL in adolescence, another issue de- The baseline testing was performed over the first
serves attention because the association between (pre- 2 weeks of the 2017/18 school year, and involved ad-
vious) sports participation and PA changes that oc- olescents who personally agreed to participate and
cur later in life (i.e., when the participants are no lon- whose parent(s) signed the consent form. Although
ger actively involved in sports) are not known. By other the testing was anonymous, the participants were in-
words, the following question arises: “Is sports partici- structed to use self-selected anonymous codes for iden-
pation earlier in adolescence related to a higher or low- tification purposes in both testing waves (i.e., the last
er PAL later in life?”. However, studies have rarely exam- 3 digits of their e-mail password). The examiners in-
ined this problem. The most likely reason for the lack of formed the participants of the study purpose and aims,
such information is the fact that answering this question that the testing was absolutely voluntary and anony-
actually requires prospective analyses. By other words, mous, and that they could leave the questionnaire form
the participants should be observed over repeated mea- or some questions unanswered. The testing lasted ap-
surements (e.g., testing waves), including the period prox. 15 min. The follow-up testing was performed
when they were actively involved in organized sports, at the end of the fourth year of high school (during
and later in life, when they were no longer involved in the last 3 weeks of the school year). The procedure and
organized sports. the study were in accordance with the Declaration of
The aim of this study was to prospectively evaluate Helsinki, and were originally approved by the Ethics
the influence of sports participation, and sociodemo- Board of the University of Split, Faculty of Kinesiology,
graphic factors on PAL and changes in PAL in adoles- Split, Croatia. After obtaining the positive opinion of
cents. The leading hypotheses of the research were that the Ethics Board, the study was approved by school au-
sports participation was positively related to PAL, and thorities as well.
negatively related to changes in PAL over the observed After the follow-up testing, the analysis of attrition
2-year period (at 16–18 years of age). bias was performed. The results showed that the re-
Nr 6 Physical activity in adolescence3

sponders (those who were tested in both testing waves) such as PA during physical education classes, sports,
and the non-responders (those who were tested at base- and active transportation. The ninth item does not
line only) did not differ in PAL at baseline (t-test = 0.14, contribute to the overall score but is used solely to
p > 0.05) or in factors related to sports participation target those participants who were not able to par-
(χ2: 0.14–1.55, p > 0.05). However, significantly more ticipate in regular PA due to illness or injury. By this
males than females dropped out between baseline and study, the authors focused on 3 variables derived by
follow-up (χ2: 16.11, p < 0.01), which was probably re- PAQ-A:
lated to the fact that boys are known to be more ab- ■ PAL at study baseline (PAL-BL),
sent from school than girls [13]. Additionally, the in- ■ PAL at follow-up (PAL-FU),
tracluster correlation (IC) for the baseline PAL (with ■ changes in PAL between baseline and follow-up
the schools as the clusters) showed good within-school (calculated as differences between PAL-BL and
variance (IC = 0.05 and 0.06 for baseline and follow-up, PAL-FU).
respectively) [14]. For the purpose of logistic regression calculation
(please see later for Statistics), the PAL-BL and PAL-FU
Variables results were dichotomized. Specifically, scores <2.73
The variables included the participants’ sociodemo- were observed as low PAL, while scores above >2.73
graphic characteristics, sports factors, and PAL. All were observed as normal PAL, as suggested previous-
variables were collected by previously validated ques- ly [17].
tionnaires [6,13,15].
Sociodemographic factors were self-reported sex, Statistics
age (in years), self-reported socioeconomic status (re- Descriptive statistics included means and standard de-
sponses included: below average, average, above av- viations (for numerical variables), and percentages
erage), maternal education and paternal educa- and frequencies (for ordinal and nominal variables).
tion (both reported on a scale with the following re- To identify the changes in PAL between baseline and
sponses: elementary school, high school, college, uni- follow-up, a t-test for dependent samples was calculat-
versity degree). Sports factors were tested at baseline ed. Differences between boys and girls in the studied
and included questions about students’ sports partic- variables were evidenced either by the t-test for an inde-
ipation [13]: pendent sample (for PAL), or by Mann-Whitney U test,
■ involvement in competitive team sports (i.e., basket- and the χ2 test (for non-parametric variables).
ball, football/soccer, handball); To calculate the associations between the studied
■ involvement in competitive individual sports variables, Spearman’s rank order correlation was ap-
(i.e., track and field, swimming, artistic gymnas- plied. To identify the associations between predictors,
tics) (both reported as “no,” “quit,” “yes”; later, and PAL-BL and PAL-FU, logistic regression was cal-
for the purpose of logistic regression calculation, culated. Previous studies identified a strong effect of sex
grouped into “non-involved” [the first 2 responses], on PAL in adolescence [6]; therefore, 2 logistic regres-
and “involved”); sion models were calculated: crude (model 0 – uncon-
■ the highest competitive sports achievement/result trolled for covariates, and model 1 – controlled for sex
(from never involved/competed, locally, to interna- as a covariate). Finally, those predictors that were found
tional level); as being significantly correlated with criteria were si-
■ the duration of sports involvement (never partici- multaneously included in (multivariate) logistic regres-
pated, <1 year, 2–5 years, >5 years). sion calculations.
The actual PAL was determined by PAQ-A. A p-value of 95% was applied, and the Statistica
Previously, PAQ-A was repeatedly found to be a valid ver. 13.0 statistical package (Statsoft, Tulsa, OK) was used
and reliable questionnaire in samples of participants for all calculations.
similar to the one observed herein [15,16]. Generally,
PAQ-A consists of 9 items asking the participants to
provide a 7-day self-reported recall. The final theo- RESULTS
retical numerical score on PAQ-A ranges from 0 (the Generally, PAL significantly decreased over the study
minimum PAL) to 5 (the maximum PAL). The first period in the total sample (2.48±0.70–2.27±0.73, t-test:
8 items ask the participants about different types of PA, 6.11, p < 0.001), among boys (2.74±0.70–2.52±0.77,
 4 D. Sekulic et al. Nr 6

¥
4
and PA-BL and PAL-FU. By general, all sports factors
PAL [score ]

¥ were correlated with PAL, with stronger correlations

* observed for PAL-BL (Spearman’s ρ: 0.22–0.44), than
* for PAL-FU (Spearman’s ρ: 0.20–0.40). The socioeco-
3
* nomic status was not correlated with PAL, but specif-
ic associations were found between paternal education
and PAL. By brief, both paternal and maternal edu-
2
cation levels were positively correlated with PAL-BL,
but only maternal education was significantly cor-
related with PAL-FU. No significant correlations be-
1
tween the studied variables and changes in PAL which
occurred between baseline and follow-up were evi-
denced (Table 2).
0
total boys girls Results of the logistic regression for the dichoto-
baseline follow-up Participants mized PAL-BL and PAL-FU criteria are presented in
Presented as M±SD. Figure 2. The male sex was a significant predictor of
Significant (p < 0.001) differences between (¥) and within groups (*). PAL-BL and PAL-FU (model 0: OR: 1.51, 95% CI: 1.23–
Figure 1. Descriptive statistics for physical activity levels (PAL) 1.87, and OR: 1.31, 95% CI: 1.01–1.61 for PAL-BL and
as obtained by the Physical Activity Questionnaire for Adolescents PAL-FU, respectively). Therefore, the authors focused
(PAQ-A) for adolescents from southern Croatia (late 2016)
on the results of model 1 (calculated with “sex” as a co-
variate). Maternal education (OR: 1.38, 95% CI: 1.02–
t-test: 4.44, p < 0.001), and girls (2.16±0.56 –1.96±0.56, 1.90), and team sports participation (OR: 1.56, 95% CI:
t-test: 4.72, p < 0.001). Higher PAL was evidenced in boys 1.11–1.87) were significant predictors on PAL-BL
than in girls, at both testing waves (t-test: 9.14 and 8.26 (Figure 2). Similarly, higher maternal-education
for PAL-BL and PAL-FU, respectively) (Figure 1). (OR: 1.35, 95% CI: 1.08–1.70), and involvement in team
Boys and girls differed significantly in all sports fac- sports (OR: 1.45, 95% CI: 1.01–1.90) were significant
tors, with boys being more involved both in individual predictors of PAL-FU.
and team sports (Mann-Whitney U test: 5.83 and 7.26, The multivariate logistic regression, with maternal
p < 0.01, respectively). Also, boys achieved better com- education and team sports participation as predictors,
petitive achievement, and were involved in sports for showed a significant association between team sports
a longer time than girls (Mann-Whitney U test: 8.31 and PAL-BL (OR: 1.50, 95% CI: 1.09–1.99). Meanwhile,
and 4.29, p < 0.01, respectively) (Table 1). team sports (OR: 1.44, 95% CI: 1.01–1.98), and mater-
Spearman’s correlation evidenced significant posi- nal education (OR: 1.40, 95% CI: 1.02–1.88) were sig-
tive associations between baseline sports participation, nificant multivariate predictors of PAL-FU.

Table 1. Descriptive statistics for the studied sociodemographic and sports factors for adolescents from southern Croatia (late 2016)
with differences between boys and girls

Participants
[n]
(N = 411)
Variable Z
boys girls
total
(N = 223) (N = 184)
F % F % F %
Socioeconomic status 0.47

below average 6 1.5 5 2.2 1 0.5

average 382 92.9 201 90.1 177 96.2

above average 23 5.6 17 7.6 6 3.3

missing 0 0.0 0 0.0 0 0.0

Nr 6 Physical activity in adolescence5

Table 1. Descriptive statistics for the studied sociodemographic and sports factors for adolescents from southern Croatia (late 2016)
with differences between boys and girls – cont.
Participants
[n]
(N = 411)
Variable Z
boys girls
total
(N = 223) (N = 184)
F % F % F %
Paternal education 1.49
elementary school 26 6.3 8 3.6 18 9.8
high school 299 72.7 164 73.5 131 71.2
college degree 48 11.7 28 12.6 20 10.9
university degree 38 9.2 23 10.3 15 8.2
missing 0 0.0 0 0.0 0 0.0
Maternal education 1.48
elementary school 115 27.9 53 23.7 61 33.2
high school 246 59.9 142 63.7 101 54.9
college degree 25 6.1 14 6.3 11 6.0
university degree 25 6.1 14 6.3 11 6.0
missing 0 0.0 0 0.0 0 0.0
Participation in sports
individual –5.83***
yes, currently 88 21.4 71 31.8 16 8.7
quit 137 33.3 77 34.5 57 31.0
no, never 186 45.3 75 33.6 111 60.3
missing 0 0.0 0 0.0 0 0.0
team –7.26***
yes, currently 94 22.9 74 33.2 19 10.3
quit 169 41.1 104 46.6 63 34.2
no, never 148 36.0 45 20.2 102 55.4
missing 0 0.0 0 0.0 0 0.0
Experience in sports 8.31***
never participated 107 26.0 27 12.1 80 43.5
<1 year 91 22.1 42 18.8 48 26.1
2–5 years 113 27.5 74 33.2 38 20.7
>5 years 100 24.3 80 35.9 18 9.8
missing 0 0.0 0 0.0 0 0.0
Sports competitive achievement 4.19***
never involved/competed 212 51.6 90 40.4 122 66.3
local 168 40.9 115 51.6 49 26.6
national level 31 6.6 15 6.7 12 6.5
international level 4 1.0 3 1.3 1 0.5
missing 0 0.0 0 0.0 0 0.0

Z – Mann-Whitney U test.
*** p < 0.001.
6 D. Sekulic et al. Nr 6

Table 2. Spearman’s rank correlation coefficients between the studied sociodemographic and sports factors, and physical activity (PA)
indices in the total sample, and separately for boys and girls (adolescents from southern Croatia, late 2016)

Spearman’s correlation
Variable
PA at baseline PA at follow-up PA difference
Socioeconomic status
total 0.02 –0.03 0.05
boys –0.02 –0.05 0.05
girls 0.00 –0.04 0.05
Paternal education
total 0.18*** 0.11* 0.07
boys 0.15* 0.07 0.07
girls 0.16* 0.09 0.07
Maternal education
total 0.21*** 0.19*** 0.06
boys 0.19* 0.21** 0.05
girls 0.23** 0.24** 0.07
Participation in individual sports
total –0.37*** –0.30*** –0.02
boys –0.26*** –0.21** –0.02
girls –0.33*** –0.24** –0.04
Participation in team sports
total –0.44*** –0.38*** –0.01
boys –0.37*** –0.32*** –0.02
girls –0.28*** –0.23** 0.01
Experience in sports
total 0.41*** 0.40*** –0.01
boys 0.35*** 0.30*** 0.03
girls 0.25*** 0.24** –0.05
Sports competitive achievement
total 0.33*** 0.33*** –0.03
boys 0.33*** 0.31*** 0.00
girls 0.22** 0.20** –0.03

* p < 0.05, ** p < 0.01, *** p < 0.001.

DISCUSSION ing the results directly related to the study aims, the dif-
ferences between sexes, and the changes in PAL that oc-
The results of this study revealed several important curred in the studied period of 2 years, are overviewed.
findings with regard to the study aims. First, the stud- Although the differences between sexes were not
ied sports participation and sociodemographic factors the direct study aim, the fact that boys had higher
were significantly correlated with PAL at baseline and PAL than girls deserves attention. Generally, studies
follow-up, with some differences between the sexes and consistently reported higher PALs among adolescent
the type of sport. Next, changes in PAL that occurred boys [3,6]. Such differences are particularly evident in
during the course of the study were not significantly late adolescence and this is mostly explained by the fact
correlated with the observed predictors. Before discuss- that boys are more involved in organized sports than
Nr 6 Physical activity in adolescence7

females. Knowing the significant differences in sports a)

factors between sexes in this study (i.e., boys were sys- Sport_Achiev
tematically more engaged in sports than girls), the sig- Sport_Exper
nificant differences in PAL between sexes were expected.
Team_Sport
The results of this study showed a significant decrease
in PAL at the age of 16–18 years, and this is in accordance Ind_Sport

with recent studies where such trends were evidenced Mat_Edu

on the country territory. By brief, Stefan et al. [18] evi- Pat_Edu
denced a PAL decrease in adolescents from the Croatian
SES
capital city, Zagreb, between the first and second grade
of high school, and this is in accordance with recent in- Male gender

vestigation done with somewhat older adolescents from 0 1 2 3 4

southern regions of the country [3]. Although in this model 0 model 1 OR (95% CI)
study there were some numerical differences between
the sexes (a somewhat larger decrease in PAL among b)
boys), the negative trends in PAL observed in the study Sport_Achiev
are supportive to previous reports where authors ob- Sport_Exper
served similar negative trends in PAL globally.
Team_Sport
For example, U.S. authors observed a substantial de-
crease in moderate to vigorous PA in adolescence, par- Ind_Sport

ticularly for girls (decreasing 5.9–4.9 h/week from ear- Mat_Edu

ly to mid-adolescence, and 5.1–3.5 h/week from mid- to Pat_Edu
late adolescence). This was accompanied by an increase
SES
in free-time computer use, particularly among boys (in-
creasing 11.4–15.2 h/week from early to mid-adoles- Male gender
cence, and 10.4–14.2 h/week from mid- to late adoles- 0 1 2 3 4
cence) [19]. Additionally, Norwegian authors showed model 0 model 1 OR (95% CI)
an evident decrease in PAL from early adolescence Ind_Sport – current participation in individual sports, Mat_Edu – maternal education
(approx. 13 years of age) to early adulthood (23 years of level, Pat_Edu – paternal education level, SES – socioeconomic status, Sport_Achiev –
sports competitive achievement, Sport_Exper – experience (engagement) in sports,
age) [20]. Finally, the results of a recent study performed Team sports – current participation in team sports.
in adolescents from Bosnia and Herzegovina showed Model 0 – non-controlled for covariates, model 1 – controlled for sex as a covariate.
similar negative trends among boys and girls that were Figure 2. Results of the logistic regression calculation – correlates
prospectively observed in late adolescence [6]. of physical activity levels at a) baseline and b) follow-up among
adolescents from southern Croatia
Sports participation plays an important role in PAL
in childhood and adolescence [21]. Indeed, a large
proportion of the PAL variance in youth is explained and PAL (both measured at baseline and follow-up)
by participation in organized sports and the influ- were expected [9–11]. Meanwhile, it is important to
ence of sports involvement on the total PAL increas- highlight that this study demonstrated a stronger in-
es [10]. The main explanation for the increasing influ- fluence of team sports participation on PAL, which is
ence of sports participation on PAL in adolescence is a novel finding, to some extent. The explanation for
related to a corresponding decrease in everyday PA due such a result may be found in the specific sociocul-
to reduced active transportation (i.e., walking, cycling), tural context of team sports participation in the stud-
and the lack of physical work (as a result of mechaniza- ied region (southern Croatia). Namely, the most popu-
tion) [22]. Taking into account the prevailing trends in lar team sports in the studied region (i.e., football/soc-
the modern society, the involvement in sports has be- cer, basketball, handball) are frequently played even out
come an even more important factor influencing PAL. of the context of organized sports, and it is not rare for
Knowing the results of previous studies where au- children and adolescents to engage in such activities out
thors regularly observed higher PAL among athlet- of a “competitive environment,” just for fun.
ic youth than among non-athletic youth, the results of Recent studies have confirmed that while sports
a consistent positive correlation between sports factors participation increases PAL directly, children who are
8 D. Sekulic et al. Nr 6

actively involved in sports are also more likely to have of sports participation in adolescence against a decrease
higher levels of everyday PA than their non-athletic in PAL. From the authors’ perspective, as former ath-
peers [23]. The mechanism of such influence was ex- letes and active sport pedagogues, the most likely expla-
plained through differences in the fitness status be- nation of such findings could be found in the “orienta-
tween sports participants and non-exercising children. tion to success” philosophy, which is the dominant ap-
By brief, those children who were engaged in sports proach to youth sports in the country. By other words,
were more likely to have better fitness, allowing them sports programs in adolescence are mostly oriented to-
to be more physically active overall [23]. Putting these wards competitive achievement; therefore, children are
findings in the context of this study and the previous- not properly educated about the additional positive out-
ly stated results of a stronger association between team comes of sports participation that will be important lat-
sports participation and PAL, the results can be dis- er in life (i.e., health-related fitness, positive social con-
cussed in further detail. It seems reasonable that not sequences).
only the fitness status (which is logically more devel- While self-perceived socioeconomic status was
oped in children who practice organized sports) but al- not found to be related to PAL, it is important to note
so “physical literacy” and “familiarity” with most pop- the specific influence of parental education on PAL and
ular sports activities plays an important role in overall changes in PAL. Since this study lacked some important
PAL. Namely, it is more likely that those children and information that will hopefully help in a more elabo-
adolescents who feel competent to participate in sports rated discussion of these findings (i.e., the authors have
will be engaged in them, even in their free time (out no information about the real socioeconomic status of
of organized sports trainings). By this study, participa- the families), the focus is solely on those interpreta-
tion in team sports may have been observed as being tions that could be supported by other study findings.
more favorable to physical literacy simply because team Paternal influence on sports participation is known to
sports are played more often than individual sports, be strong, especially for boys [24]. Therefore, the re-
even in free time. sults of a positive association between paternal edu-
The correlations between sports factors and PAL cation and PAL should also be observed as expected.
were, to some extent, influenced by sex (note that However, although the associations between maternal
some ORs for some sports factors did not reach statis- education and PAL were not so frequently confirmed
tical significance when sex was included as a covariate in literature, recent studies have reported a specific as-
in model 1). Taking into account the significant influ- sociation between these indices. Specifically, in a previ-
ence of the male sex on PAL-BL (for the crude logis- ously cited study performed among adolescents from
tic regression model, exclusively), it is logical to con- Bosnia and Herzegovina, the authors reported a posi-
clude that the identified associations between sports tive influence of maternal education on changes in PAL
factors and PAL were more evident in boys. This is un- among girls [6]. Therefore, the fact that mothers have
derstandable knowing the differences in sports partici- a strong influence on their children’s behavior in late ad-
pation between boys and girls. Specifically, studies have olescence can still be supported.
regularly confirmed that boys are more engaged in or- In Croatia, mothers are generally known to be “more
ganized sports than girls, and this was directly observed responsible” for children’s educational achievement than
in the wider territory of former Yugoslavia (note that fathers. Actually, these “educational specifics” are not
Croatia is one of the countries which was part of former characteristic only for the studied country but rather are
Yugoslavia) [13,24,25]. Finally, this is evidenced herein global paradigms [27,28]. Therefore, the background of
also. Therefore, it is logical that sex is a strong factor of the influence of maternal education on changes in PAL
influence on PAL. should be observed as a result of the strong influence
These results did not confirm that sports participa- of mothers on children’s educational achievement [29].
tion at the age of 16 years was a factor that influences Participation in organized competitive sports decreases
changes in PAL at 16–18 years of age. This is a important in late adolescence. Thus, the PAL in later adolescence is
finding since this period of life is known to be character- more likely to be a consequence of “personal awareness
ized by a significant decrease in PAL, while sports par- of the health-related benefits of PAL than was the case
ticipation is frequently observed as “preventive” against in earlier adolescence (when children participated in
such a decrease [26]. Therefore, these results actually do sports because of the fun and/or as a result of parental
not confirm the hypothesis about the “protective effect” control). Therefore, the positive influence of maternal
Nr 6 Physical activity in adolescence9

education on PA changes in this period of life should be vation points to the specific maternal influence on
observed as a consequence of the positive influence of health-related knowledge among adolescents, which
maternal education on the overall health-related behav- should be explored more specifically in the future.
ior of children [29]. By doing so, special attention should be placed on
the evaluation of other important health-related issues
Limitations and strengths in adolescents (i.e., quality of nutrition, sexual health).
The fact that the variables evaluated were self-report-
ed is the most significant limitation of the study, espe-
cially for some variables were self-report bias is partic- ACKNOWLEDGMENTS
ularly possible (i.e., socioeconomic status of the fami- The authors are particularly grateful to all children who vol-
ly). Additionally, the fact that the study did not observe untary participated in the study. Special thanks go to the
all theoretically important determinants of PAL, such school authorities who approved the investigation and assist-
as educational variables (i.e., educational achievement), ed with the research.
and other types of activity that could be theoretically
important covariates of the established relationships
(i.e., the availability of a playground, parental PAL), REFERENCES
cannot be ignored. Finally, this study observed only sev- 1. Pavičić Žeželj S, Kenđel Jovanović G, Krešić G. The asso-
eral determinants of the PAL-changes; therefore, fur- ciation between the Mediterranean diet and high phys-
ther studies are needed in order to explore the problem ical activity among the working population in Croatia.
more specifically. Med Pr. 2019;70(2):169–76, https://doi.org/10.13075/
This is one of few studies in which predictors of mp.5893.00773.
changes in PAL in late adolescence were determined and, 2. Śmigielski J, Ruszkowska J, Piotrowski W, Polakowska M,
to the best of the authors’ knowledge, this is the first in- Bielecki W, Hanke W, et al. The relationship between phys-
vestigation of its kind in Croatia. Additionally, the study ical activity level and selected cardiovascular risk fac-
indicated various sports factors as potential correlates tors and mortality of males ≥ 50 years in Poland – the re-
of PAL, which is another important strength of the in- sults of follow-up of participants of National Multicenter
vestigation. Therefore, the authors hope that their re- Health Survey WOBASZ. Int J Occup Med Environ He­
sults will contribute to the knowledge in the field and alth. 2016;29(4):633–48, https://doi.org/10.13075/ijomeh.
initiate further research. 1896.00660.
3. Sekulic D, Blazevic M, Gilic B, Kvesic I, Zenic N. Prospective
Analysis of Levels and Correlates of Physical Activity
CONCLUSIONS During COVID-19 Pandemic and Imposed Rules of Social
Generally, PAL significantly decreased between the age Distancing; Gender Specific Study Among Adolescents
of 16–18 years, and such negative trends were evident in from Southern Croatia. Sustainability. 2020;12(10):4072,
both sexes. Therefore, it is of utmost importance to de- https://doi.org/10.3390/su12104072.
velop methods and policies that will influence such neg- 4. Best K, Ball K, Zarnowiecki D, Stanley R, Dollman J.
ative trends in Croatian adolescents. By Search of Consistent Predictors of Children’s Physical
Sports participation was indicated as an import- Activity. Int J Environ Res Public Health. 2017;14(10):1258,
ant determinant of PAL in late adolescence, with high- https://doi.org/10.3390/ijerph14101258.
er PAL among athletic adolescents. Additionally, par- 5. Kleszczewska D, Mazur J, Siedlecka J. Family, school and
ticipation in team sports was found to be more influ- neighborhood factors moderating the relationship between
ential than participation in individual sports, which physical activity and some aspects of mental health in ado-
is explained by sociocultural background (practicing lescents. Int J Occup Med Environ Health. 2019;32(4):423–
team sports is possible even in free time), and the im- 39, https://doi.org/10.13075/ijomeh.1896.01389.
portance of overall physical literacy in achieving higher 6. Miljanovic Damjanovic V, Obradovic Salcin L, Zenic N,
PAL. However, the results of the study did not confirm Foretic N, Liposek S. Identifying Predictors of Changes
the influence of sports participation on changes in PAL in Physical Activity Level in Adolescence: A Prospective
at 16–18 years of age. Analysis in Bosnia and Herzegovina. Int J Environ Res
Parental education positively influenced PAL, with Public Health. 2019;16(14):2573, https://doi.org/10.3390/
higher influence of maternal education. This obser- ijerph16142573.
10 D. Sekulic et al. Nr 6

7. Mann S, Beedie C, Jimenez A. Differential effects of aer- 17. Benítez-Porres J, Alvero-Cruz JR, Sardinha LB, López-
obic exercise, resistance training and combined exercise Fernández I, Carnero EA. Cut-off values for classify-
modalities on cholesterol and the lipid profile: review, syn- ing active children and adolescentes using the Physical
thesis and recommendations. Sports Med. 2014;44:211– Activity Questionnaire: PAQ-C and PAQ-ACut-off val-
21, https://doi.org/10.1007/s40279-013-0110-5. ues for classifying active children and adolescents using
8. Van Der Horst K, Paw MJ, Twisk JW, Van Mechelen W. the Physical Activity Questionnaire: PAQ-C and PAQ-A.
A brief review on correlates of physical activity and seden- Nutr Hosp. 2016;33(5):564, https://doi.org/10.20960/
tariness in youth. Med Sci Sports Exerc. 2007;39(8):1241– nh.564.
50, https://doi.org/10.1249/mss.0b013e318059bf35. 18. Stefan L, Soric M, Devrnja A, Petric V, Misigoj-
9. Sprengeler O, Buck C, Hebestreit A, Wirsik N, Ahrens W. Durakovic M. One-year changes in physical activity and
Sports Contribute to Total Moderate to Vigorous Physical sedentary behavior among adolescents: the Croatian
Activity in School Children. Med Sci Sports Exerc. 2019;51(8): Physical Activity in Adolescence Longitudinal Study
1653–61, https://doi.org/10.1249/MSS.0000000000001948. (CRO-PALS). Int J Adolesc Med Health. 2018, https://doi.
10. Vella SA, Schranz NK, Davern M, Hardy LL, Hills AP, org/10.1515/ijamh-2017-0223.
Morgan PJ, et al. The contribution of organised sports 19. Nelson MC, Neumark-Stzainer D, Hannan PJ, Sirard JR,
to physical activity in Australia: Results and directions Story M. Longitudinal and secular trends in phys-
from the Active Healthy Kids Australia 2014 Report Card ical activity and sedentary behavior during adoles-
on physical activity for children and young people. J Sci cence. Pediatrics. 2006;118(6):e1627–34, https://doi.org/
Med Sport. 2016;19(5):407–12, https://doi.org/10.1016/ 10.1542/peds.2006-0926.
j.jsams.2015.04.011. 20. Kjønniksen L, Torsheim T, Wold B. Tracking of lei-
11. Kobel S, Kettner S, Kesztyüs D, Erkelenz N, Drenowatz C, sure-time physical activity during adolescence and young
Steinacker JM. Correlates of habitual physical activi- adulthood: a 10-year longitudinal study. Int J Behav Nutr
ty and organized sports in German primary school chil- Phys Act. 2008;5:69, https://doi.org/10.1186/1479-5868-
dren. Public Health. 2015;129(3):237–43, https://doi.org/ 5-69.
10.1016/j.puhe.2014.12.002. 21. Sallis JF, Prochaska JJ, Taylor WC. A review of correlates
12. Miller J, Pereira M, Wolfson J, Laska M, Nelson T, of physical activity of children and adolescents. Med
Neumark-Sztainer D. Developmental Trends and De­ter­ Sci Sports Exerc. 2000;32(5):963–75, https://doi.org/10.
mi­nants of Physical Activity From Adolescence to Adult­ 1097/00005768-200005000-00014.
hood Differ by Ethnicity/Race and Sex. J Phys Act He­ 22. Booth VM, Rowlands AV, Dollman J. Physical activity
alth. 2018;15(5):345–54, https://doi.org/10.1123/jpah. temporal trends among children and adolescents. J Sci
2017-0287. Med Sport. 2015;18(4):418–25, https://doi.org/10.1016/
13. Zenic N, Terzic A, Ostojic L, Sisic, N, Saavedra JM, j.jsams.2014.06.002.
Kristjansdottir H, et al. Educational and sport factors as 23. Telford RM, Telford RD, Cochrane T, Cunningham RB,
predictors of harmful alcohol drinking in adolescence: Olive LS, Davey R. The influence of sport club partic-
a prospective study in Bosnia and Herzegovina. Int J Public ipation on physical activity, fitness and body fat during
Health. 2019;64(2):185–94, https://doi.org/10.1007/s00038- childhood and adolescence: The LOOK Longitudinal
018-1168-5. Study. J Sci Med Sport. 2016;19(5):400–6, https://doi.org/
14. Killip S, Mahfoud Z, Pearce K. What is an intracluster cor- 10.1016/j.jsams.2015.04.008.
relation coefficient? Crucial concepts for primary care re- 24. Tahiraj E, Cubela M, Ostojic L, Rodek J, Zenic N,
searchers. Ann Fam Med. 2004;2(3):204–8, https://doi. Sekulic D, et al. Prevalence and Factors Associated with
org/10.1370/afm.141. Substance Use and Misuse among Kosovar Adolescents;
15. Samaržija DV, Mišigoj-Duraković M. [Reliability of Cross Sectional Study of Scholastic, Familial-, and
Croatian Version of the Questionnaire for Assessment of Sports-Related Factors of Influence. Int J Environ Res
Overall Level of Physical Activity]. Hrvatski Šport­sko­me­ Public Health. 2016;13(5):502, https://doi.org/10.3390/
di­cinski Vjesnik. 2013;28(1):24–32. Croatian. ijerph13050502.
16. Pojskic H, Eslami B. Relationship Between Obesity, Physical 25. Idrizovic K, Zenic N, Tahirajl E, Rausavljevic N, Sekulic D.
Activity, and Cardiorespiratory Fitness Levels in Children Cigarette Smoking Among 17–18 Year Old Adolescents –
and Adolescents in Bosnia and Herzegovina: An Analysis Prevalence and Association With Sociodemographic, Familial,
of Gender Differences. Front Physiol. 2018;9:1734, https:// Sport, and Scholastic. Factors. Med Pr. 2015;66(2):153–63,
doi.org/10.3389/fphys.2018.01734. https://doi.org/10.13075/mp.5893.00104.
Nr 6 Physical activity in adolescence11

26. Dumith SC, Gigante DP, Domingues MR, Kohl HW 3rd. 28. Grolnick WS, Ryan RM. Parent styles associated with chil-
Physical activity change during adolescence: a systematic dren’s self-regulation and competence in school. J Educ
review and a pooled analysis. Int J Epidemiol. 2011;40(3): Psychol. 1989;81(2):143.
685–98, https://doi.org/10.1093/ije/dyq272. 29. Fan X, Chen M. Parental involvement and students’ aca-
27. Lewko JH, Ewing ME. Sex differences and parental in- demic achievement: A meta-analysis. Educ Psychol Rev.
fluence in sport involvement of children. J Sport Psychol. 2001;13(1):1–22.
1980;2(1):62–8.

This work is available in Open Access model and licensed under a Creative Commons Attribution-NonCommercial 3.0 Poland License – http://creative-
commons.org/licenses/by-nc/3.0/pl/deed.en.
Publisher: Nofer Institute of Occupational Medicine, Łódź, Poland

View publication stats