Journal Logo

BASIC SCIENCES: Epidemiology

Emotional and Behavioral Problems in Relation to Physical Activity in Youth


Author Information
Medicine & Science in Sports & Exercise: October 2008 - Volume 40 - Issue 10 - p 1749-1756
doi: 10.1249/MSS.0b013e31817b8e82
  • Free


Existing evidence from several large-scale population surveys points to relatively widespread presence of mental health problems among adolescents (10,28,29). It has been estimated that more than 30% of Americans aged 9 to 17 yr have a diagnosable mental health disorder that causes some or significant functional impairment (27). According to a recent national survey in Finland, 13% of adolescents had symptoms of moderate to serious depression (15).

The beneficial effects of regular physical activity on mental health are fairly well documented in the adult population (3). Clinical and epidemiologic studies, not only on depression and anxiety (7, 17,18) but also on panic attacks (7,17,18), phobias (7,18), and stress disorders (18), have shown inverse associations between physical activity and these disorders. In adults, a dose-response relation between physical activity and mental disorders has also been reported (6,7).

Much less is known about the association between physical activity and mental health problems in adolescents. Large cross-sectional studies have suggested physical activity level being inversely associated with emotional and behavioral problems among young people (1,11,23). In follow-up studies, it has also been reported that natural change occurring in physical activity across time during adolescent years is inversely associated with a change in depressive symptoms (14) and that there is a consistent relationship between physical activity in adolescence and psychological well-being in adulthood (19).

Previous studies notwithstanding the literature about physical activity and mental health problems among young people are scarce (5,14,17). Especially, cohort studies with the general population are few (14). Most of the earlier studies have focused on emotional problems, primarily depression and anxiety, but only a few have examined behavioral or social problems and even fewer have viewed the phenomenon in a broad sense, including both internalizing and externalizing syndromes (5,17). Many previous studies have measured emotional and behavioral problems using a method originally developed for adults without establishing the validity among adolescents (14).

To enhance the mental well-being of youth, we need more information on factors associated with emotional and behavioral problems. The present study explores how the level of physical activity is associated with self-reported emotional and behavioral problems among adolescents. We hypothesize that a low level of moderate- to vigorous- intensity physical activity (MVPA) is associated with increased prevalence of internalizing syndromes, externalizing syndromes, social problems, thought problems, and attention problems in a general population of adolescents.


Study setting and population.

The study population of this cross-sectional study comprised the Northern Finland Birth Cohort 1986 that originally contained 9432 children who were born alive and whose expected date of birth was between July 1, 1985 and June 30, 1986 in the two northernmost provinces of Finland, namely, Oulu and Lapland (8). In 2001-2002, at the age of 15 to 16 yr (hereafter called "16 yr"), all those 9215 cohort members who were alive and whose addresses were known were sent a postal questionnaire including questions about their physical activity and their emotional and behavioral problems (response rate = 80%, N = 7344). Their parents were also sent a questionnaire that included questions on family income, mother and father's education, and family structure (response rate = 76%, N = 6985). The present cross-sectional analyses were conducted in 2007 including 7002 subjects who completed the Youth Self-Report (YSR) questionnaire assessing their emotional and behavioral problems. Informed consent was obtained from all participants and their parents, and the research protocol was approved by the ethics committee of the University Hospital of Oulu.

Emotional and behavioral problems.

Emotional and behavioral problems at the age of 16 yr were measured using the YSR that is a widely used questionnaire designed to assess emotional and behavioral problems among adolescents aged 11 to 18 yr (2). The YSR was scored for the following eight syndrome scales: 1) anxious/depressed symptoms, 2) withdrawn/depressed symptoms, 3) somatic complaints, 4) social problems, 5) thought problems, 6) attention problems, 7) rule-breaking behavior, and 8) aggressive behavior. The response alternatives and the scores for these statements were either not true (score = 0), somewhat or sometimes true (score = 1), or very true or often true (score = 2) from the preceding 6 months.

The item scores were summed up to obtain a summary score for each subscale, the maximum subscale score being 26 for anxious/depressed symptoms, 16 for withdrawn/depressed symptoms, 20 for somatic complaints, 22 for social problems, 24 for thought problems, 18 for attention problems, 30 for rule-breaking behavior, and 34 for aggressive behavior. The original summary scores were trichotomised into normal range, borderline range, and clinical range groups based on the recommended cutoff points (approximately the 84th through 90th percentiles) (2). For further analyses borderline range and clinical range were combined into one group and is hereafter called "problem range." Syndrome scales 1-3 comprise the internalizing scale, scales 4-6 are not part of either internalizing or externalizing scales and are called other syndromes in this article, whereas scales 7 and 8 comprise the externalizing scale.

The reliability and the validity of the YSR have been documented by Achenbach et al. (2). The content validity of the YSR problem item scores has been supported by four decades of research, consultation, feedback, and revision and by findings that all items discriminated significantly (P < 0.01) between demographically matched referred and nonreferred children. The criterion-related validity of the YSR scales was supported by multiple regressions, odds ratios (OR), and discriminant analyses, all of which showed significant (P < 0.01) discrimination between referred and nonreferred children. The construct validity of the scales has been supported by evidence for significant association with analogous scales of other instruments and with the DSM criteria, by cross-cultural replications of the syndromes, by genetic and biochemical findings, and by predictions of long-term outcomes.

Physical activity.

At 16 yr, the subjects were asked about their participation in moderate- to vigorous-intensity physical activity (MVPA) outside school hours. The amount of MVPA was evaluated by asking, "How much do you participate in brisk physical activity outside school hours?" In the questionnaire, the term "brisk" was defined as a physical activity causing at least some sweating and getting out of breath. The response alternatives were not at all, approximately 0.5 h·wk−1, 1 h·wk−1, 2-3 h·wk−1, 4-6 h·wk−1, and ≥7 h·wk−1. Adolescents were classified into three groups according to their weekly participation in MVPA: 1) active (≥4 h of MVPA), 2) moderately active (2-3 h of MVPA), and 3) inactive (≤1 h of MVPA). The test-retest reliability of this question has been reported to be fairly good, with an intraclass correlation coefficient being 0.83 among Finnish adolescents aged 15 to 16 yr (24).

Potential confounding factors.

Family type, annual family income, parents' level of education, and children's obesity level were considered as potential confounding factors in the analyses. These factors have earlier been described in detail (9). Family type was divided into four categories: 1) always a two-parent family (76% of adolescents), 2) single-parent family (13%), 3) reconstructed family (10%), and 4) always a one-parent family (1%). Annual income was enquired in the parents' questionnaire, and household consumption units were calculated by taking into account household size and structure (9,16,26). The families were classified into quartiles based on their annual income per consumption unit: 1) less than U20AC;9200, 2) U20AC;9200-13,599, 3) U20AC;13,600-18,599, and 4) more than U20AC;18,600 per year. Mother and father's highest level of education was also categorized according to educational level categories used by the International Standard Classification of Education (22) and the Finnish National Board of Education (25). Four educational groups were formed: 1) basic education, lasting ≤9 yr (mothers = 9%, fathers = 15%); 2) upper secondary education, lasting 10-12 yr (66%, 67%); 3) tertiary education, lasting ≥13 yr (13%, 12%); and 4) other or degree not finished (11%, 7%) (9). Body weight and height were both self-reported in the postal inquiry and measured in the health examination at 15-16 yr. Self-reported body weight and height were used for those who failed to attend the health examination. Body mass index (BMI) was calculated as weight divided by the square of their height (kg·m−2). Overweight was defined as BMI between the 85th and 95th percentiles (24.20-28.16 kg·m−2 in boys and 24.13-27.70 kg·m−2 in girls) and obesity as BMI above the 95th percentile (≥28.18 kg·m−2 in boys and ≥27.70 kg·m−2 in girls) (30).

Missing values.

Individuals were excluded from the analyses if YSR questionnaire data were missing for more than eight items (not counting open-ended and socially desirable items, altogether 16 items) and if there were more than one answer missing on any of the eight subscales. Otherwise, the missing values were replaced by the mean value of the items on that particular scale for that individual.

Statistical analyses.

ANOVA with LSD post hoc analysis was used to compare the mean scores of eight different syndromes for active, moderately active, and inactive group. Cross-tabulation and multivariable logistic regression were used to evaluate how the level of MVPA associated with the prevalence of emotional and behavioral problems. Odds ratios (OR) and their 95% confidence intervals (95% CI) for having different syndromes were calculated by the level of MVPA and adjusted for family type, annual income, parents' level of education, and BMI. Statistical analyses were carried out using SPSS software, version 14.0 (21).


The proportion of girls scoring on the problem range was substantially higher for all eight syndrome scales compared with boys (Table 1). In boys, the prevalence of emotional and behavioral problems varied between 1.6% and 8.6%, thought problems being the rarest and rule-breaking behavior being the most common syndrome. Similarly in girls, thought problems were the rarest (4.1%) and rule-breaking behavior was the most common (15.6%) syndrome. Boys were physically more active than girls (P < 0.001); 46% of boys and 29% of girls were classified as physically active, 24% of boys and 30% of girls were moderately active, and 31% of boys and 41% of girls were physically inactive.

The prevalence of emotional and behavioral problems in boys (n = 3355) and girls (n = 3647) aged 15 to 16 yr.

Mean scores of the different syndromes are presented according to the levels of MVPA in Figures 1 and 2. Among the boys, the mean scores of anxious/depression, withdrawal/depression, social and attention problems, and rule-breaking behavior were significantly higher for the inactive group compared to the active group (Fig. 1), indicating that physically inactive individuals have more emotional and social problems than physically active individuals. Also, in the moderately active group, the mean scores of anxious/depression, withdrawal/depression, and social problems were significantly higher than in the active group. Similarly in girls, the mean scores of anxious/depression, withdrawal/depression, social and attention problems, rule-breaking behavior, and aggressive behavior were highest in the physically inactive group. However, in girls, the mean syndrome scores of the moderately active group did not differ significantly from the mean scores of the active group (Fig. 2).

Mean syndrome scores with 95% CI in boys (n = 3307) by the level of MVPA at age 16 yr. ANOVA/LSD post hoc significance compared to the active group: *P < 0.05.
Mean syndrome scores with 95% CI in girls (n = 3601) by the level of MVPA at age 16 yr. ANOVA/LSD post hoc significance compared to the active group: *P < 0.05.

With regard to internalizing syndromes, physically inactive adolescents had the highest prevalence of anxious/depressed symptoms in girls and withdrawn/depressed symptoms as well as somatic complaints in both boys and girls (Table 2). The adjusted results from regression analyses showed that in boys being moderately active (OR = 3.21) or inactive (OR = 2.93) was significantly associated with anxious/depressed symptoms compared with those being physically active (Table 3). Similarly, inactivity was associated with withdrawn/depressed problems in both boys and girls (Table 3). In girls, physical inactivity was also associated with somatic complaints (OR = 1.40).

The prevalence (%) of emotional and behavioral problems according to participation in MVPA in boys (N = 3355) and girls (n = 3647).
Logistic regression of the internalizing syndromesa by the level of physical activity in boys (n = 2078) and girls (n = 2134).

Social and attention problems were most common among physically inactive boys and girls (Table 2). After adjustments, in boys, being moderately active (OR = 2.43) and inactive (OR = 3.59) was associated with social problems compared to being active. In girls, physical inactivity was associated with social problems (OR = 3.18). Physical inactivity was associated with thought problems in boys (OR = 2.34) but not in girls (Table 4). Physical inactivity was associated with attention problems in both boys (OR = 1.87) and girls (OR = 2.06).

Logistic regression of the other syndromesa by the level of physical activity in boys (n = 2078) and girls (n = 2134).

With regard to externalizing syndromes, the prevalence of rule-breaking behavior was highest among physically inactive boys and girls (Table 2). After controlling for potential confounding factors, physical inactivity was associated with rule-breaking behavior in girls (OR = 1.82), but in boys, this association was not significant (OR = 1.33; Table 5). Physical inactivity was not associated with aggressive behavior in either boys or girls (Table 5). In girls, being moderately active was associated with the lowest prevalence of aggressive behavior (OR = 0.53) compared with being physically active.

Logistic regression of the externalizing syndromesa by the level of physical activity in boys (n = 2078) and girls (n = 2134).


In this study, physical inactivity in adolescents was associated with an increased likelihood of having several emotional and behavioral problems: anxious/depressed symptoms in boys, withdrawn/depressed symptoms in boys and girls, somatic complaints in girls, rule-breaking behavior in girls, social and attention problems in boys and girls, and thought problems in boys.

The present results are consistent with earlier findings from similar population-based cross-sectional studies, which suggest physical activity being inversely associated with depression (1,11,14), anxiety (11), somatic complaints (23), and social problems (11) in youth. However, the present study is the first one to give an extensive picture of the association of mental health problems and physical activity including both internalizing and externalizing mental health syndromes.

There is a growing evidence in adults that an increase in physical activity is associated with a reduction of depressive and anxiety symptoms, suggesting both psychological and physiological mechanisms to explain the beneficial effects of exercise on mental health (6,7,17,18). In their recent review on physical activity and mental health, Paluska and Schwenk (17) successfully contend the psychological mechanism under four theories: the distraction hypothesis, the self-efficacy theory, the mastery hypothesis, and the social interaction hypothesis. The proposed theories behind the physiological mechanisms are the monoamine hypothesis, the endorphin hypothesis, and the thermogenic model.

These results give a reason to believe that similar associations and background mechanisms might exist in adolescents. The inverse relation between physical activity and rule-breaking behavior could, to some extent, be explained by the emotional and educational influence of physical activities; playing sports both in organized and informal settings can be an excellent way of discharging feelings, and it also teaches cooperativeness and complying with rules. Improvement in teamwork, tolerance, and self-directedness through many types of physical activities might explain why active adolescents have fewer social problems than their inactive counterparts. Reduced attention problems in physically active adolescents compared to inactive adolescents might be due to better concentration and perception, abilities that are exercised through various types of physical activity. However, in this cross-sectional study setting, it is not fertile to refer to any specific mechanism described above. It might be that an integrative psychobiological model that combines components of each hypothesis offers the most likely explanation for the beneficial effects of physical activity on mental health. More well-designed studies are needed for achieving a clear consensus regarding the previously proposed mechanisms (17).

Results of the present study provide some evidence of the possible threshold effect of physical activity on mental health problems in youth. It seems that in boys, being moderately active does not provide positive mental health effects, but higher level of activity is needed. In boys, this seems most evident according to internalizing syndromes and social problems. Instead, in girls it might be that even a moderate level of MVPA has benefits on mental health, especially on externalizing syndromes. These results highlight the gender variability in the possible threshold effect of MVPA on mental health problems and further investigation of the relevance of current physical activity guidelines for mental health in adolescence (4).

In this article, the definition of "active" is not strictly consistent with current physical activity recommendations suggesting that school-aged children should participate in MVPA for at least 1 h·d−1. This level was met by only 23%of boys and 10% of girls in our population (24), resulting in too small groups for statistical analyses, especially in girls. In addition, we did not want to choose extremely active adolescents as the reference group in the analyses. Therefore, we named those (46% of boys and 29% of girls) who participated in MVPA for ≥4 h·wk−1 as active. The inactive group participated in MVPA for ≤1h·wk−1, which is absolutely too little to have optimal health benefits at this age.

However, the cross-sectional setting in the present study does not allow us to make strong conclusions about causality between physical activity and emotional and behavioral problems. It might be that some of the adolescents who reported less physical activity did so because of mental health problems. For example, adolescents with anxious/depressed or withdrawn/depressed problems may be less likely to engage in physically strenuous activities because of low energy or apathy (18). On the other hand, adolescents with higher levels of problem behavior may not receive as much support and encouragement for their participation, for example, in team games as well-behaved adolescents, and may therefore be more inclined to drop out or choose not to be involved in the first place.

The results of the present study form a basis for future research needed to investigate the possible causality between physical activity and mental health problems and the mechanisms behind the observed association. Because of the widespread presence of mental health problems and concerns about the clinical effectiveness and side effects of psychological, psychosocial, and biological treatments in adolescents, low-risk and low-cost interventions are needed alongside prevailing therapeutic practices and medical treatment. A few studies have so far evaluated exercise in the prevention and treatment of anxiety and depression among children and young people, but the effect of exercise is unknown because the evidence base is scarce (12). However, physical activity may play an important role as a relatively simple, low-risk, and cost-effective therapy for mental health problems. In addition, because a physically inactive lifestyle seems to be more common among adolescents with mental health problems, encouraging them to engage in physical activities is important to prevent other detrimental health effects of a sedentary lifestyle on cardiometabolic health, for instance.


The main strengths of the present study include the use of a large general population of adolescents as a study group and the measurement of mental health problems in a broad sense including eight syndromes of emotional and behavioral problems. On the other hand, as previously noted, the cross-sectional design of the present study does not permit an examination of the causal relationship between physical activity and mental health problems. The results of this study are based on the data from self-reports of both physical activity and mental health problems and may thus be more prone to measurement errors than more objective measurements. Social desirability bias may lead to overestimating the time or intensity of physical activity, and for children and youth, errors in recall are likely to be greater than for adults (13,20). It is also possible that some chronic medical conditions or their treatments not measured in this study have a confounding effect on the association between physical activity and mental health problems.


Physical inactivity was associated with several emotional and behavioral problems in a general population of adolescent boys and girls. Widespread presence of emotional and behavioral problems in adolescence and the relatively stable nature of physical activity pattern from youth to adulthood make this age group particularly important to address with future studies and physical activity interventions. Future research is encouraged to investigate the mechanisms behind these associations and the effectiveness of physical activity in the treatment of emotional and behavioral problems among young people.

This study was partly funded by the Ministry of Education, Finland.


1. Abu-Omar K, Rütten A, Lehtinen V. Mental health and physical activity in the European Union. Soz Praventivmed. 2004;49:301-9.
2. Achenbach TM, Rescorla LA. Manual for the ASEBA School Age Forms & Profiles. Burlington (VT): University of Vermont, Research Center for Children, Youth, & Families; 2001. 238 p.
3. Biddle JH, Mutrie N. Psychology of Physical Activity: Determinants, well-being and interventions. New York (NY): Routledge; 2001. p. 167-253.
4. Desha LN, Ziviani JM, Nicholson JM, Martin G, Darnell RE. Physical activity and depressive symptoms in American adolescents. J Sport Exerc Psychol. 2007;29:534-43.
5. Donaldson SJ, Ronan KR. The effects of sports participation on young adolescents' emotional well-being. Adolescence. 2006;41:369-89.
6. Dunn AL, Trivedi MH, O'Neal HA. Physical activity dose-response effects on outcomes of depression and anxiety. Med Sci Sports Exerc. 2001;33(6 suppl):S587-97.
7. Goodwin RD. Association between physical activity and mental disorders among adults in the United States. Prev Med. 2003;36:698-703.
8. Järvelin M-R, Hartikainen-Sorri A-L, Rantakallio P. Labour induction policy in hospitals of different levels of specialisation. Br J Obstet Gynaecol. 1993;100:310-5.
9. Kantomaa MT, Tammelin TH, Näyhä S, Taanila AM. Adolescents' physical activity in relation to family income and parents' education. Prev Med. 2007;44:410-5.
10. Kessler RC, Tat Chiu W, Demler O, Walters EE. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62:617-27.
11. Kirkcaldy BD, Shephard RJ, Siefen RG. The relationship between physical activity and self-image and problem behaviour among adolescents. Soc Psychiatry Psychiatr Epidemiol. 2002;37:544-50.
12. Larun L, Nordheim LV, Ekeland E, Hagen KB, Heian F. Exercise in prevention and treatment of anxiety and depression among children and young people. Cochrane Database Syst Rev [Internet]. 2006 [cited 2007 August 3]; 3. Available from: Art. No.: CD004691. doi: 10.1002/14651858.CD004691.pub2.
13. Montoye HJ, Kemper HCG, Saris WHM, Washburn RA. Measuring Physical Activity and Energy Expenditure. Champaign (IL): Human Kinetics; 1996. p. 42-62.
14. Motl RW, Birnbaum AS, Kubik MY, Dishman RK. Naturally occurring changes in physical activity are inversely related to depressive symptoms during early adolescence. Psychosom Med. 2004;66:336-42.
15. National Research and Development Centre for Welfare and Health Web site [Internet]. Finland: National Research and Development Centre for Welfare and Health (STAKES): School Health Research 2004; [cited 2007 August 3]. Available from:
16. OECD. The OECD List of Social Indicators. Paris (France): OECD; 1982. 124 p.
17. Paluska A, Schwenk TL. Physical activity and mental health: current concepts. Sports Med. 2000;29:167-80.
18. Richardson CR, Faulkner G, McDevitt J, Skrinar GS, Hutchinson DS, Piette JD. Integrating physical activity into mental health services for persons with serious mental illness. Psychiatr Serv. 2005;56:324-31.
19. Sacker A, Cable N. Do adolescent leisure-time physical activities foster health and well-being in adulthood? Evidence from two British birth cohorts. Eur J Public Health. 2006;16:331-5.
20. Shephard RJ. Limits to the measurement of habitual physical activity by questionnaires. Br J Sports Med. 2003;37:197-206.
21. SPSS Base 14.0 User's Guide. Chicago (IL): SPSS Inc., 2005. p.339-47.
22. Statistics Finland. UNESCO's International Standard Classification of Education ISCED 1997: Finnish Implementation Manual 2000: Appendix 1 to The Finnish Standard Classification of Education 2000. Helsinki (Finland): Statistics Finland; 2001. 169 p.
23. Steptoe A, Butler N. Sports participation and emotional wellbeing in adolescents. Lancet. 1996;347:1789-92.
24. Tammelin T, Ekelund U, Remes J, Näyhä S. Physical activity and sedentary behaviors among Finnish youth. Med Sci Sports Exerc. 2007;39(7):1067-74.
25. The Finnish National Board of Education Web site [Internet]. Finland: The Finnish National Board of Education: The Education System of Finland; [cited 2007 August 3]. Available from:,4699.
26. United Nations. Provisional Guidelines on Statistics of the Distribution of Income, Consumption and Accumulation of Households. New York (NY): United Nations; 1977. 97 p.
27. US Department of Health and Human Services. Mental Health: A Report of the Surgeon General. Rockville (MD): US Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Mental Health Services, National Institutes of Health, National Institute of Mental Health [Internet]. 1999 [cited 2007 August 3]. Available from:
28. Verhulst FC, van der Ende J, Ferdinand RF, Kasius MC. The prevalence of DSM-II-R-diagnoses in a national sample of Dutch adolescents. Arch Gen Psychiatry. 1997;54:329-36.
29. Volleberg WAM, van Dorsselaer S, Monshouwer K, van der Ende J, ter Bogt T. Mental health problems in early adolescents in the Netherlands. Soc Psychiatry Psychiatr Epidemiol. 2006;41:156-63.
30. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med. 1997;337:869-73.


©2008The American College of Sports Medicine