Association between physical exercise and psychosocial problems in 96 617 Dutch adolescents in secondary education: a cross-sectional study

Association between physical exercise and psychosocial problems in 96 617 Dutch adolescents in... Abstract Background Psychosocial problems negatively affect school performance, social skills and mental development. In recent years, researchers have investigated the relationship between physical activity and psychological health. With this large school-based study, we examined whether physically inactive adolescents and slightly active adolescents experience more psychosocial problems compared with active adolescents. Methods This study is based on the Dutch National Youth Health Monitor. This monitor uses a, school-based, cross-sectional questionnaire conducted among 96 617 adolescents in 2015. To examine the association between physical exercise and psychosocial problems, multi-level linear regression was carried out. Results The weighted average Strengths and Difficulties Questionnaire score of active adolescents was lower than that of inactive adolescents. Adolescents who are inactive had 12% (β = 1.12; 95% CI: 1.10–1.14; P <0 .001) more psychosocial problems compared with active adolescents. Further, inactive adolescents had a higher score on the subscales emotional problems (β = 1.19; 95% CI: 1.17—1.22; P < 0.001) and problems with peers (β = 1.16; 95% CI: 1.14—1.19; P < 0.001). There was no statistical significant difference in total score of the Strengths and Difficulties Questionnaire between active and slightly active adolescents. Conclusion Physically active adolescents have fewer psychosocial problems compared with physically inactive adolescents. Not only is this association significant, but there is an indication that it is also of clinical relevance. Introduction According to the population-based studies, 20% of the Dutch adolescents have some degree of psychosocial problems.1,2 Psychosocial problems can be a psychological problem as well as a social problem. These adolescents may have anxious feelings and thoughts and also may have problems in social interaction with others.3 Psychosocial problems among adolescents negatively affect school performance, social skills and mental development. Moreover, they lead to high social costs and are a major cause of increased health care use in later life.4–7 Researchers have investigated the relationship between physical activity and psychological health. Several cross-sectional studies have shown that regular exercise is associated with a lower risk of psychosocial problems in adults and children.8–16 In a large-scale population study, however, no positive association was found between physical exercise and psychological health after controlling for several confounders.17 Various longitudinal studies have shown that people who exercise regularly are less likely to develop symptoms of anxiety or depression.8,18–23 In contrast, some longitudinal studies found no relationship between exercise and mental health.24–26 The focus of these studies is on the association between physical exercise and psychological health instead of psychosocial problems. The youth health care in the Netherlands focuses on psychosocial problems in children. Encouraging adolescents to perform more physical activity might be an effective intervention to prevent psychosocial problems. Using data of a large school-based study, we examined whether physically inactive adolescents and slightly active adolescents experience more psychosocial problems compared with active adolescents. Methods Study design This study is based on a secondary data analysis of the National Dutch Youth Health Monitor 2015 from the National Institute for Public Health and Environment and 25 Municipal Public Health Service (MPHS). This monitor is a school-based cross-sectional questionnaire conducted among adolescents in the second and fourth grade of regular secondary education. The Dutch Youth Health Monitor is conducted every 4 year by all 25 MPHS offices in the Netherlands. Recruitment and study population The sampling frame consisted of all regular secondary schools in the Netherlands enlisted with the ministry of education. A stratified sampling procedure was used. A proportional number of schools by educational level was randomly selected (N = 567). In total, 376 of the 567 contacted schools enrolled (66%). The schools that did not participate were too busy or had no interest. The Dutch Youth Health Monitor included adolescents in the second and fourth grade at regular Dutch secondary schools. A total of 96 617 adolescents completed the online self-report questionnaire during a regular lesson at school in the period of October–December 2015. Measures The monitor consists of 75 questions (189 items). The following are examples of topics were included in the questionnaire: grade, age, level of education, sex, ethnicity, family composition, living environment, perceived health, perceived happiness, psychosocial health, height and weight, physical exercise, nutrition, smoking, alcohol, drugs, school experience, bullying, social media, gaming and sexuality. The Dutch Youth Health Monitor includes validated instruments such as the Strengths and Difficulties Questionnaire (SDQ) and several questions about health and lifestyle. The Dutch version of the SDQ was used to measure the degree of psychosocial problems.27 The SDQ consists of 25 items with three possible answers: not true, somewhat true and certainly true. The SDQ is divided into five subscales: emotional problems, conduct problems, hyperactivity, peer problems and pro-social behavior. The total SDQ score is a fully dimensional measure: an increase in the total SDQ score corresponds with an increase in the risk of mental health disorder.28 Compared with the other scales, an opposite interpretation applies for the pro-social behavior scale. Hence, a higher score means more pro-social behavior. Widenfelt et al.29 and Muris et al.30 examined psychometric properties of the Dutch translation of the self-reported SDQ in adolescents. They concluded that the psychometric properties of the self-reported Dutch translation of the SDQ are acceptable. In 2017, the Dutch National Health council concluded on the basis of recent scientific insights that no advice can be given about the amount of exercise in relation to the positive effects on health of children. Therefore, the Dutch National Health council still advises to maintain the recommendation of at least 1 h of moderate intensive exercise every day.31 In this study, physical exercise was measured with a single-item question: ‘For how many days per week do you practice a sport or engage in physical exercise for at least one hour? Include all kinds of sports or exercise you do in a day.’ Response choices were: (Almost) Never, 1 day per week, 2 days per week, 3 days per week, 4 days per week, 5 days per week, 6 days per week and every day. Adolescents were considered active, according to the Dutch standard of healthy exercise, by answering ‘every day’ to this question.32 In this study, adolescents were considered inactive if they exercised <3 days a week. With regard to the respondents’ psychosocial health, we compared the respondents who answered ‘every day’ (active), as a reference group, with those who answered ‘3–6 days per week’ (slightly active) and with those who answered ‘<3 days a week’ (inactive). Probable sources of confounding were identified in the literature.10,14,19,26,33,34 The following covariates and related categories (between brackets) were included in the analysis: gender, grade (2 or 4), age(12–13, 14–15, 16–17) household situation (parents living together vs. other answers), ethnicity (non-immigrant, non-Western, Western immigrant), urbanization (highly, moderately, not urbanized), educational level (low, intermediate, high), use of soft drugs last month (no/yes), lifetime use of hard drugs (no/yes), lifetime use of soft drugs (no/yes), alcohol or binge drinking previous four weeks (no/yes) and being bullied in the past 3 months (no/yes). These covariates are divided into three categories: demographic, socialeconomic and lifestyle covariates. Statistical methods To examine the adjusted associations between physical exercise and psychosocial problems, multi-level linear regression was carried out with random effects to account for clustering of adolescents within schools. The total SDQ score and the categorical variable physical exercise were included in the model. A second multi-level model was fitted for the adjusted association between physical exercise and psychosocial problems with a fixed effect for the following demographic covariates: gender, grade, age, household situation, ethnicity and urbanization. A third multi-level model was fitted for the above-mentioned association with fixed effects for the demographic covariates and the following socioeconomic covariate: educational level. A fourth multi-level model was fitted with fixed effects for demographic, socioeconomic and the following lifestyle covariates: lifetime use of hard drugs, use of soft drugs and alcohol last month, binge drinking and being bullied. These four models were also fitted in a multi-level model to determine the association between the five sub-scale scores of the SDQ and physical exercise. The logarithmically transformed regression coefficients of the total SDQ score corrected for demographic, socioeconomic and lifestyle covariates appeared to be most relevant in this study. The relative differences between the regression coefficients in total SDQ score of inactive and active adolescents presented in this model were considerable compared with the other models. Therefore, both adjusted and unadjusted models are presented in table 3. All analyses were conducted in MlwIN V.2.22 and SPSS 20. The normality of the primary outcome measures was visually inspected. The outcome variables were log transformed because they were positively skewed. The distribution of the total SDQ score ranges from 0 to 40. It is not possible to logarithmically transform zeros. For that reason we added a constant to each answer on the 25 items [log(x + 1)] to make all results positive and not zero.35,36 No imputation technique was used. The fractions of missing values of the outcome variables and independent variables were very small (<0.5%). The influence of missing values on the association between physical exercise and psychosocial problems was considered to be very low. We used a database of 96 617 adolescents. Due to this size, a very small effect size could be statistically significant.37 Therefore, a P values smaller than 0.001 was considered statistically significant. Results Most adolescents in our sample were 14 or 15 years old (44%), living with both their parents (76%), had a non-immigrant background (81%) and had a low educational level (52%). Almost a third of the adolescents had consumed alcohol in the previous month (32%) and more than a fifth had used alcohol excessively in the previous 4 weeks (23%). Few adolescents had ever used hard drugs (4%) or had used soft drugs in the previous 4 weeks (5%). As shown in table 1, more girls than boys were inactive (19% vs. 15%). Other noteworthy differences in this sample were that older adolescents were more often inactive than younger adolescents (23% vs. 14%). Furthermore, adolescents who have a non-Western background and adolescents who were attending a lower level of education were more often inactive, compared with non-immigrant adolescents (27% vs. 15%) and adolescents who attend a higher level of education (21% vs. 10%). Finally, adolescents who had ever used hard drugs or had used soft drugs in the previous 4 weeks were more likely to be inactive compared with adolescents who had never used hard drugs (21% vs. 16%) or used soft drugs in the past 4 weeks (23% vs. 16%). Table 1 Adolescents demographic, socio-economic and lifestyle characteristics in relation to physical exercise levels, N (%)   Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572    Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572  Table 1 Adolescents demographic, socio-economic and lifestyle characteristics in relation to physical exercise levels, N (%)   Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572    Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572  The weighted average total SDQ score of active adolescents [9.33 ± 5.05 (SD)] was lower that of the inactive adolescents (10.97 ± 5.45; see table 2). There is no difference between the active and slightly active adolescents (9.30 ± 4.83). Table 2 Total SDQ and subscale scores in relation to physical exercise levels, mean (SD)   Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)    Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)  Table 2 Total SDQ and subscale scores in relation to physical exercise levels, mean (SD)   Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)    Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)  The logarithmically transformed coefficients of the unadjusted and adjusted models that estimate the association between the frequency of physical exercise and psychosocial problems are presented in table 3. Both unadjusted as adjusted coefficients between active and inactive adolescents are significant. Table 3 Multi-level linear regression results of continuous (logarithmically transformed) mean of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use, being bullied. * P < 0.001 Table 3 Multi-level linear regression results of continuous (logarithmically transformed) mean of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use, being bullied. * P < 0.001 After the antilog procedure, the average total SDQ score of adolescents who were inactive was 1.12 times higher than that of adolescents who were active (β = 1.12; 95% CI: 1.10–1.14; P < 0.001; see table 4). In other words, inactive adolescents were likely to have psychosocial problems compared with active adolescents. Table 4 Multi-level linear regression results of continuous mean after the antilog procedure of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use and being bullied. * P < 0.001. Table 4 Multi-level linear regression results of continuous mean after the antilog procedure of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use and being bullied. * P < 0.001. Additionally, it was relatively more common for inactive adolescents to have emotional problems (β = 1.19; 95% CI: 1.17–1.22; P < 0.001), conduct problems (β = 1.03; 95% CI: 1.01–1.05; P < 0.001), hyperactivity (β = 1.03; 95% CI: 1.00–1.05; P < 0.001), problems with peers (β = 1.16; 95% CI: 1.14–1.19; P < 0.001) and exhibit less pro-social behavior (β = 0.94; 95% CI: 0.93–0.95; P < 0.001) compared with active adolescents. The association between the average SDQ score of active adolescents and that of slightly active adolescents was almost statistically significant according to our threshold of 0.001 (β = 1.01; 95% CI: 1.00–1.03; P = 0.0012). Emotional problems (β = 1.08; 95% CI: 1.06–1.09; P < 0.001) were more common in slightly active adolescents compared with active adolescents. The weak but statistically significant association of conduct problems and problems with peers between these groups: 0.97 (95% CI: 0.96–0.98; P < 0.001) and 1.02 (95% CI: 1.01–1.04; P < 0.001), respectively. Discussion The aim of this study was to explore whether inactive and slightly active adolescents experience more psychosocial problems compared with active adolescents. Main findings of this study Our data provide evidence of an inverse association between the physical activity level of adolescents in the Netherlands and the extent of their psychosocial problems. Inactive adolescents experience more psychosocial problems than active adolescents. Similarly, inactive adolescents experience more emotional and conduct problems, hyperactivity, problems with peers and show less pro-social behavior. The differences between slightly active and active adolescents are less distinctive. For example, there is no statistical significant difference in total SDQ score between active and slightly active adolescents. Clinical relevance also plays a role in the interpretation of the observed association. Beforehand, we considered a result clinically relevant if the difference in adjusted regression (β) coefficient between active and inactive adolescents was greater than one point. Depending on the baseline value, an increase in one point on the SDQ scale could imply that an adolescent will be referred to a psychologist by the MPHS. The difference in weighted average total SDQ score between active and inactive adolescents is more than 1.5 point (table 2). This indicates, to our standards, a clinically relevant difference in psychosocial problems. However, due to the skewness of the distribution, we had to logarithmically transform the outcome measures. Therefore, the regression coefficients have to be interpreted as a ratio between those groups of adolescents instead of a difference. As a consequence, clinical relevance could not been deduced from these regression coefficients. What is already know on this topic The above-mentioned findings corroborates with earlier cross-sectional findings.11,13,14,16 These studies showed that lack of exercise was associated with various psychological problems. Psychosocial problems consist partly of psychological problems, making these findings relevant. However, these results are not in line with some longitudinal studies.24–26 Furthermore, Stavrakakis et al. focused on the onset of depressive disorders and physical activity but found no evidence of such an effect of physical activity on depression.22,38 This implies that the association we found in our data might be of limited value. On the other hand, the same Stavrakakis also studied the prospective relationship between physical activity and depressive symptoms in adolescents and found a weak statistically significant negative association.22,39 This shows that the relationship between physical activity and psychological health is complex. Genetic and biological factors, personality traits and environmental factors such as socialeconomic status and lifestyle habits interact and jointly affect psychosocial problems.22,38 In combination with these factors, sufficient physical exercise might prevent or mitigate psychosocial problems. What this study adds It was known in advance that most studies showed that physical exercise were associated with psychological problems. However, none of these studies focused on psychosocial problems. Psychosocial health is the outcome measure of the SDQ, widely used in youth care. This research has shown that physical exercise is also associated with less of psychosocial health. Finally, we used the data of the nationwide Dutch Youth Health Monitor conducted among almost 100 000 adolescents. Because of this number of participants, the results presented in this research are, as far as we know, unique. Limitations of this study The first weakness of this study is the cross-sectional design, which implies that a causal direction cannot be determined. Therefore, it may be that a lack of exercise causes psychosocial problems, but it is also possible that psychosocial problems lead to less exercise. Finally, it is also possible for an underlying factor to affect both physical exercises as psychosocial problems. The direction of causation can be established in a randomized controlled trial or in a prospective design. Second, the amount of physical activity was determined with a single-item question. This question covered the number of days per week of moderately intensive exercise for at least 1 h. Additional information about the exact amount and intensity was not collected. Information about the type of exercise, such as team sport or an individual activity, could deepen our understanding. Conclusion This study shows that physically active adolescents have fewer psychosocial problems than inactive adolescents. This association is not only statistically significant, but we think that it is also of clinical relevance. Inactive adolescents experience more emotional and conduct problems, hyperactivity, problems with peers and they show less pro-social behavior. Compared with active adolescents, slightly active adolescents experienced more emotional problems and problems with peers. However, slightly active adolescents experienced fewer conduct problems compared with active adolescents. Acknowledgements The data set of the Youth Health Monitor 2015 was provided by the Dutch National Institute for Public Health and Environment and 25 Municipal Public Health Service. Conflicts of interest: None declared. Key points Previous studies have demonstrated that 20% of the Dutch adolescents have some degree of psychosocial problems. Previous research investigated the relationship between physical activity and psychological health. However, none of these studies focused on psychosocial problems. We used the data of the nationwide Dutch Youth Health Monitor conducted among almost 100 000 adolescents. This study shows that physically active adolescents have fewer psychosocial problems than inactive adolescents. The amount of physical activity was determined with a single-item question. Information about the manner of exercise could deepen our understanding about the association between physical activity and psychosocial health. References 1 Theunissen MHC, Vogels AGC, Reijneveld SA. Work experience and style explain variation among pediatricians in the detection of children with psychosocial problems. Acad Pediatr  2012; 12: 495– 501. Google Scholar CrossRef Search ADS PubMed  2 Crone MR, Bekkema N, Wiefferink CH, et al.   Professional identification of psychosocial problems among children from ethnic minority groups: room for improvement. J Pediatr  2010; 156: 277– 84 e1. Google Scholar CrossRef Search ADS PubMed  3 Vannieuwenborg L, Buntinx F, De Lepeleire J. Presenting prevalence and management of psychosocial problems in primary care in Flanders. Arch Public Health  2015; 73: 10. Google Scholar CrossRef Search ADS PubMed  4 Anthony LG, Anthony BJ, Glanville DN, et al.   The relationships between parenting stress, parenting behaviour and preschoolers' social competence and behaviour problems in the classroom. Infant Child Dev  2005; 14: 133– 54. Google Scholar CrossRef Search ADS   5 Steinberg L, Lamborn SD, Dornbusch SM, et al.   Impact of parenting practices on adolescent achievement: authoritative parenting, school involvement, and encouragement to succeed. Child Dev  1992; 63: 1266– 81. Google Scholar CrossRef Search ADS PubMed  6 Rushton J, Bruckman D, Kelleher K. Primary care referral of children with psychosocial problems. Arch Pediatr Adolesc Med  2002; 156: 592– 8. Google Scholar CrossRef Search ADS PubMed  7 Smith JP, Smith GC. Long-term economic costs of psychological problems during childhood. Soc Sci Med  2010; 71: 110– 5. Google Scholar CrossRef Search ADS PubMed  8 Ten Have M, De Graaf R, Monshouwer K. [ Sports and Psychological Health. Results of the Netherlands Health Survey and Incidence Study (NEMESIS) ]. Utrecht: Trimbos Institute, 2009. 9 De Moor MH, Beem AL, Stubbe JH, et al.   Regular exercise, anxiety, depression and personality: a population-based study. Prev Med  2006; 42: 273– 9. Google Scholar CrossRef Search ADS PubMed  10 Goodwin RD. Association between physical activity and mental disorders among adults in the United States. Prev Med  2003; 36: 698– 703. Google Scholar CrossRef Search ADS PubMed  11 Kirkcaldy BD, Shephard RJ, Siefen RG. The relationship between physical activity and self-image and problem behaviour among adolescents. Soc Psych Psych Epidemol  2002; 37: 544– 50. Google Scholar CrossRef Search ADS   12 Muhsen K, Lipsitz J, Garty-Sandalon N, et al.   Correlates of generalized anxiety disorder: independent of co-morbidity with depression: findings from the first Israeli National Health Interview Survey (2003-2004). Soc Psych Psych Epidemiol  2008; 43: 898– 904. Google Scholar CrossRef Search ADS   13 Steptoe A, Wardle J, Fuller R, et al.   Leisure-time physical exercise: prevalence, attitudinal correlates, and behavioral correlates among young Europeans from 21 countries. Prev Med  1997; 26: 845– 54. Google Scholar CrossRef Search ADS PubMed  14 Tao FB, Xu ML, Kim SD, et al.   Physical activity might not be the protective factor for health risk behaviours and psychopathological symptoms in adolescents. J Paediatr Child Health  2007; 43: 762– 7. Google Scholar CrossRef Search ADS PubMed  15 Taylor MK, Pietrobon R, Pan D, et al.   Healthy People 2010 Physical activity guidelines and psychological symptoms: exercise from a large nationwide database. J Phys Act Health  2004; 1: 114– 30. Google Scholar CrossRef Search ADS   16 Ussher MH, Owen CG, Cook DG, et al.   The relationship between physical activity, sedentary behaviour and psychological wellbeing among adolescents. Soc Psychiatry Psychiatr Epidemiol  2007; 42: 851– 6. Google Scholar CrossRef Search ADS PubMed  17 Allison KR, Adlaf EM, Irving HM, et al.   Relationship of vigorous physical activity to psychologic distress among adolescents. J Adolescent Health  2005; 37: 164– 6. Google Scholar CrossRef Search ADS   18 Beard JR, Heathcote K, Brooks R, et al.   Predictors of mental disorders and their outcome in a community based cohort. Soc Psychiatry Psychiatr Epidemiol  2007; 42: 623– 30. Google Scholar CrossRef Search ADS PubMed  19 Brown WJ, Ford JH, Burton NW, et al.   Prospective study of physical activity and depressive symptoms in middle-aged women. Am J Prev Med  2005; 29: 265– 72. Google Scholar CrossRef Search ADS PubMed  20 Sanchez-Villegas A, Ara I, Guillen-Grima F, et al.   Physical activity, sedentary index, and mental disorders in the SUN cohort study. Med Sci Sports Exerc  2008; 40: 827– 34. Google Scholar CrossRef Search ADS PubMed  21 Strohle A, Hofler M, Pfister H, et al.   Physical activity and prevalence and incidence of mental disorders in adolescents and young adults. Psychol Med  2007; 37: 1657– 66. Google Scholar CrossRef Search ADS PubMed  22 McMahon EM, Corcoran P, O'Regan G, et al.   Physical activity in European adolescents and associations with anxiety, depression and well-being. Eur Child Adolesc Psychiatry  2016. 23 McPhie ML, Rawana JS. The effect of physical activity on depression in adolescence and emerging adulthood: a growth-curve analysis. J Adolesc  2015; 40: 83– 92. Google Scholar CrossRef Search ADS PubMed  24 De Moor MH, Boomsma DI, Stubbe JH, et al.   Testing causality in the association between regular exercise and symptoms of anxiety and depression. Arch Gen Psychiatry  2008; 65: 897– 905. Google Scholar CrossRef Search ADS PubMed  25 Kritz-Silverstein D, Barrett-Connor E, Corbeau C. Cross-sectional and prospective study of exercise and depressed mood in the elderly: the Rancho Bernardo study. Am J Epidemiol  2001; 153: 596– 603. Google Scholar CrossRef Search ADS PubMed  26 Wiles NJ, Jones GT, Haase AM, et al.   Physical activity and emotional problems amongst adolescents: a longitudinal study. Soc Psychiatry Psychiatr Epidemiol  2008; 43: 765– 72. Google Scholar CrossRef Search ADS PubMed  27 Goodman R. The Strengths and Difficulties Questionnaire: a research note. J Child Psychol Psychiatry  1997; 38: 581– 6. Google Scholar CrossRef Search ADS PubMed  28 Goodman A, Goodman R. Strengths and difficulties questionnaire as a dimensional measure of child mental health. J Am Acad Child Adolesc Psychiatry  2009; 48: 400– 3. Google Scholar CrossRef Search ADS PubMed  29 van Widenfelt BM, Goedhart AW, Treffers PD, et al.   Dutch version of the Strengths and Difficulties Questionnaire (SDQ). Eur Child Adolesc Psychiatry  2003; 12: 281– 9. Google Scholar CrossRef Search ADS PubMed  30 Muris P, Meesters C, van den Berg F. The Strengths and Difficulties Questionnaire (SDQ)–further evidence for its reliability and validity in a community sample of Dutch children and adolescents. Eur Child Adolesc Psychiatry  2003; 12: 1– 8. Google Scholar CrossRef Search ADS PubMed  31 Dutch Health Council. Guidelines for exercise 2017. 2017, Publication no. 2017/08. 32 Kemper HGC, Ooijendijk WTM, Stiggelbout M. Consensus on the Dutch Standard for Healthy Exercise. Ned Tijdschr Geneeskd  2000; 78: 180– 3. 33 Feng XQ, Wilson A. Do neighbourhood socioeconomic circumstances not matter for weight status among Australian men? Multilevel evidence from a household survey of 14 691 adults. BMJ Open  2015; 5: e007052. Google Scholar CrossRef Search ADS PubMed  34 Koezuka N, Koo M, Allison KR, et al.   The relationship between sedentary activities and physical inactivity among adolescents: results from the Canadian Community Health Survey. J Adolesc Health  2006; 39: 515– 22. Google Scholar CrossRef Search ADS PubMed  35 Chang BH, Pocock S. Analyzing data with clumping at zero. An example demonstration. J Clin Epidemiol  2000; 53: 1036– 43. Google Scholar CrossRef Search ADS PubMed  36 Lachin JM, McGee PL, Greenbaum CJ, et al.   Sample size requirements for studies of treatment effects on beta-cell function in newly diagnosed type 1 diabetes. PLoS One  2011; 6: e26471. Google Scholar CrossRef Search ADS PubMed  37 Schlomer BJ, Copp HL. Secondary data analysis of large data sets in urology: successes and errors to avoid. J Urol  2014; 191: 587– 96. Google Scholar CrossRef Search ADS PubMed  38 Stavrakakis N, Roest AM, Verhulst F, et al.   Physical activity and onset of depression in adolescents: a prospective study in the general population cohort TRAILS. J Psychiatr Res  2013; 47: 1304– 8. Google Scholar CrossRef Search ADS PubMed  39 Stavrakakis N, de Jonge P, Ormel J, et al.   Bidirectional prospective associations between physical activity and depressive symptoms. The TRAILS Study. J Adolesc Health  2012; 50: 503– 8. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The European Journal of Public Health Oxford University Press

Association between physical exercise and psychosocial problems in 96 617 Dutch adolescents in secondary education: a cross-sectional study

Loading next page...
 
/lp/ou_press/association-between-physical-exercise-and-psychosocial-problems-in-96-Sjdd7cxN0y
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.
ISSN
1101-1262
eISSN
1464-360X
D.O.I.
10.1093/eurpub/ckx230
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Psychosocial problems negatively affect school performance, social skills and mental development. In recent years, researchers have investigated the relationship between physical activity and psychological health. With this large school-based study, we examined whether physically inactive adolescents and slightly active adolescents experience more psychosocial problems compared with active adolescents. Methods This study is based on the Dutch National Youth Health Monitor. This monitor uses a, school-based, cross-sectional questionnaire conducted among 96 617 adolescents in 2015. To examine the association between physical exercise and psychosocial problems, multi-level linear regression was carried out. Results The weighted average Strengths and Difficulties Questionnaire score of active adolescents was lower than that of inactive adolescents. Adolescents who are inactive had 12% (β = 1.12; 95% CI: 1.10–1.14; P <0 .001) more psychosocial problems compared with active adolescents. Further, inactive adolescents had a higher score on the subscales emotional problems (β = 1.19; 95% CI: 1.17—1.22; P < 0.001) and problems with peers (β = 1.16; 95% CI: 1.14—1.19; P < 0.001). There was no statistical significant difference in total score of the Strengths and Difficulties Questionnaire between active and slightly active adolescents. Conclusion Physically active adolescents have fewer psychosocial problems compared with physically inactive adolescents. Not only is this association significant, but there is an indication that it is also of clinical relevance. Introduction According to the population-based studies, 20% of the Dutch adolescents have some degree of psychosocial problems.1,2 Psychosocial problems can be a psychological problem as well as a social problem. These adolescents may have anxious feelings and thoughts and also may have problems in social interaction with others.3 Psychosocial problems among adolescents negatively affect school performance, social skills and mental development. Moreover, they lead to high social costs and are a major cause of increased health care use in later life.4–7 Researchers have investigated the relationship between physical activity and psychological health. Several cross-sectional studies have shown that regular exercise is associated with a lower risk of psychosocial problems in adults and children.8–16 In a large-scale population study, however, no positive association was found between physical exercise and psychological health after controlling for several confounders.17 Various longitudinal studies have shown that people who exercise regularly are less likely to develop symptoms of anxiety or depression.8,18–23 In contrast, some longitudinal studies found no relationship between exercise and mental health.24–26 The focus of these studies is on the association between physical exercise and psychological health instead of psychosocial problems. The youth health care in the Netherlands focuses on psychosocial problems in children. Encouraging adolescents to perform more physical activity might be an effective intervention to prevent psychosocial problems. Using data of a large school-based study, we examined whether physically inactive adolescents and slightly active adolescents experience more psychosocial problems compared with active adolescents. Methods Study design This study is based on a secondary data analysis of the National Dutch Youth Health Monitor 2015 from the National Institute for Public Health and Environment and 25 Municipal Public Health Service (MPHS). This monitor is a school-based cross-sectional questionnaire conducted among adolescents in the second and fourth grade of regular secondary education. The Dutch Youth Health Monitor is conducted every 4 year by all 25 MPHS offices in the Netherlands. Recruitment and study population The sampling frame consisted of all regular secondary schools in the Netherlands enlisted with the ministry of education. A stratified sampling procedure was used. A proportional number of schools by educational level was randomly selected (N = 567). In total, 376 of the 567 contacted schools enrolled (66%). The schools that did not participate were too busy or had no interest. The Dutch Youth Health Monitor included adolescents in the second and fourth grade at regular Dutch secondary schools. A total of 96 617 adolescents completed the online self-report questionnaire during a regular lesson at school in the period of October–December 2015. Measures The monitor consists of 75 questions (189 items). The following are examples of topics were included in the questionnaire: grade, age, level of education, sex, ethnicity, family composition, living environment, perceived health, perceived happiness, psychosocial health, height and weight, physical exercise, nutrition, smoking, alcohol, drugs, school experience, bullying, social media, gaming and sexuality. The Dutch Youth Health Monitor includes validated instruments such as the Strengths and Difficulties Questionnaire (SDQ) and several questions about health and lifestyle. The Dutch version of the SDQ was used to measure the degree of psychosocial problems.27 The SDQ consists of 25 items with three possible answers: not true, somewhat true and certainly true. The SDQ is divided into five subscales: emotional problems, conduct problems, hyperactivity, peer problems and pro-social behavior. The total SDQ score is a fully dimensional measure: an increase in the total SDQ score corresponds with an increase in the risk of mental health disorder.28 Compared with the other scales, an opposite interpretation applies for the pro-social behavior scale. Hence, a higher score means more pro-social behavior. Widenfelt et al.29 and Muris et al.30 examined psychometric properties of the Dutch translation of the self-reported SDQ in adolescents. They concluded that the psychometric properties of the self-reported Dutch translation of the SDQ are acceptable. In 2017, the Dutch National Health council concluded on the basis of recent scientific insights that no advice can be given about the amount of exercise in relation to the positive effects on health of children. Therefore, the Dutch National Health council still advises to maintain the recommendation of at least 1 h of moderate intensive exercise every day.31 In this study, physical exercise was measured with a single-item question: ‘For how many days per week do you practice a sport or engage in physical exercise for at least one hour? Include all kinds of sports or exercise you do in a day.’ Response choices were: (Almost) Never, 1 day per week, 2 days per week, 3 days per week, 4 days per week, 5 days per week, 6 days per week and every day. Adolescents were considered active, according to the Dutch standard of healthy exercise, by answering ‘every day’ to this question.32 In this study, adolescents were considered inactive if they exercised <3 days a week. With regard to the respondents’ psychosocial health, we compared the respondents who answered ‘every day’ (active), as a reference group, with those who answered ‘3–6 days per week’ (slightly active) and with those who answered ‘<3 days a week’ (inactive). Probable sources of confounding were identified in the literature.10,14,19,26,33,34 The following covariates and related categories (between brackets) were included in the analysis: gender, grade (2 or 4), age(12–13, 14–15, 16–17) household situation (parents living together vs. other answers), ethnicity (non-immigrant, non-Western, Western immigrant), urbanization (highly, moderately, not urbanized), educational level (low, intermediate, high), use of soft drugs last month (no/yes), lifetime use of hard drugs (no/yes), lifetime use of soft drugs (no/yes), alcohol or binge drinking previous four weeks (no/yes) and being bullied in the past 3 months (no/yes). These covariates are divided into three categories: demographic, socialeconomic and lifestyle covariates. Statistical methods To examine the adjusted associations between physical exercise and psychosocial problems, multi-level linear regression was carried out with random effects to account for clustering of adolescents within schools. The total SDQ score and the categorical variable physical exercise were included in the model. A second multi-level model was fitted for the adjusted association between physical exercise and psychosocial problems with a fixed effect for the following demographic covariates: gender, grade, age, household situation, ethnicity and urbanization. A third multi-level model was fitted for the above-mentioned association with fixed effects for the demographic covariates and the following socioeconomic covariate: educational level. A fourth multi-level model was fitted with fixed effects for demographic, socioeconomic and the following lifestyle covariates: lifetime use of hard drugs, use of soft drugs and alcohol last month, binge drinking and being bullied. These four models were also fitted in a multi-level model to determine the association between the five sub-scale scores of the SDQ and physical exercise. The logarithmically transformed regression coefficients of the total SDQ score corrected for demographic, socioeconomic and lifestyle covariates appeared to be most relevant in this study. The relative differences between the regression coefficients in total SDQ score of inactive and active adolescents presented in this model were considerable compared with the other models. Therefore, both adjusted and unadjusted models are presented in table 3. All analyses were conducted in MlwIN V.2.22 and SPSS 20. The normality of the primary outcome measures was visually inspected. The outcome variables were log transformed because they were positively skewed. The distribution of the total SDQ score ranges from 0 to 40. It is not possible to logarithmically transform zeros. For that reason we added a constant to each answer on the 25 items [log(x + 1)] to make all results positive and not zero.35,36 No imputation technique was used. The fractions of missing values of the outcome variables and independent variables were very small (<0.5%). The influence of missing values on the association between physical exercise and psychosocial problems was considered to be very low. We used a database of 96 617 adolescents. Due to this size, a very small effect size could be statistically significant.37 Therefore, a P values smaller than 0.001 was considered statistically significant. Results Most adolescents in our sample were 14 or 15 years old (44%), living with both their parents (76%), had a non-immigrant background (81%) and had a low educational level (52%). Almost a third of the adolescents had consumed alcohol in the previous month (32%) and more than a fifth had used alcohol excessively in the previous 4 weeks (23%). Few adolescents had ever used hard drugs (4%) or had used soft drugs in the previous 4 weeks (5%). As shown in table 1, more girls than boys were inactive (19% vs. 15%). Other noteworthy differences in this sample were that older adolescents were more often inactive than younger adolescents (23% vs. 14%). Furthermore, adolescents who have a non-Western background and adolescents who were attending a lower level of education were more often inactive, compared with non-immigrant adolescents (27% vs. 15%) and adolescents who attend a higher level of education (21% vs. 10%). Finally, adolescents who had ever used hard drugs or had used soft drugs in the previous 4 weeks were more likely to be inactive compared with adolescents who had never used hard drugs (21% vs. 16%) or used soft drugs in the past 4 weeks (23% vs. 16%). Table 1 Adolescents demographic, socio-economic and lifestyle characteristics in relation to physical exercise levels, N (%)   Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572    Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572  Table 1 Adolescents demographic, socio-economic and lifestyle characteristics in relation to physical exercise levels, N (%)   Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572    Physical exercise    Active  Slightly active  Inactive  Total  Covariates  n  %  n  %  n  %  n  Gender      Male  10.206  21  31.191  64  7.006  15  48.403      Female  8.403  18  30.333  63  9.069  19  47.805  Age      12–13  7.243  20  23.918  66  4.927  14  36.088      14–15  8.024  19  27.039  64  7.106  17  42.169      16–17  3.339  19  10.561  59  4.040  23  17.940  Grade      Second class  9.791  20  31.578  65  7.481  15  48.850      Fourth class  8.818  19  29.946  63  8.594  18  47.358  Household situation      Living with two parents  14.347  20  47.582  65  10.935  15  72.864      Other answers  4.242  18  13.893  60  5.123  22  23.258  Ethnicity      Non-western immigrant  15.216  20  51.092  66  11.574  15  77.882      Western immigrants  1.149  19  3.662  60  1.236  20  6.047      Non-western immigrants  2.234  18  6.753  55  3.252  27  12.239  Urbanization      (Very) highly urbanized  7.080  18  23.722  62  7.574  20  38.376      Moderately urbanized  3.827  19  12.893  64  3.359  17  20.079      Not urbanized  7.702  20  24.909  66  5.142  14  37.753  Level of education      Low (vmbo)  9.507  19  29.969  60  10.618  21  49.821      Intermediate (havo)  5.137  19  17.794  67  3.498  13  26.429      High (vwo)  3.965  20  14.034  70  1.959  10  19.958  Use of soft drugs (previous month)      No  17.475  19  58.576  64  14.869  16  90.920      Yes  1.060  22  2.682  55  1.094  23  4.836  Hard drug use (life time)      No  17.547  19  58.916  64  15.006  16  91.469      Yes  875  23  2.195  57  806  21  3.876  Use of alcohol (previous month)      No  11.357  19  38.227  64  9.926  17  59.510      Yes  5.710  20  17.601  63  4.823  17  28.134  Binge drinking (previous month)      No  12.986  19  43.842  64  11.239  17  68.067      Yes  4.119  21  12.088  61  3.545  18  19.752  Being bullied      No  16.512  19  54.638  64  13.667  16  84.817      Yes  1.943  18  6.408  61  2.221  21  10.572  The weighted average total SDQ score of active adolescents [9.33 ± 5.05 (SD)] was lower that of the inactive adolescents (10.97 ± 5.45; see table 2). There is no difference between the active and slightly active adolescents (9.30 ± 4.83). Table 2 Total SDQ and subscale scores in relation to physical exercise levels, mean (SD)   Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)    Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)  Table 2 Total SDQ and subscale scores in relation to physical exercise levels, mean (SD)   Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)    Scale  Inactive  Slightly active  Active  Total    Range  Mean  (SD)  Mean  (SD)  Mean  (SD)  Mean  (SD)  Total difficulties score  0–40  10.97  (5.45)  9.30  (4.83)  9.33  (5.05)  9.59  (5.02)  Emotional problems  0–10  2.87  (2.29)  2.33  (2.09)  2.08  (2.04)  2.37  (2.13)  Conduct problems  0–10  1.86  (1.63)  1.54  (1.38)  1.70  (1.54)  1.63  (1.46)  Hyperactivity  0–10  4.28  (2.40)  4.02  (2.36)  4.16  (2.48)  4.09  (2.39)  Peer problems  0–10  1.97  (1.65)  1.41  (1.44)  1.39  (1.48)  1.50  (1.50)  Pro-social behavior  0–10  7.54  (1.91)  7.92  (1.68)  8.00  (1.75)  7.87  (1.74)  The logarithmically transformed coefficients of the unadjusted and adjusted models that estimate the association between the frequency of physical exercise and psychosocial problems are presented in table 3. Both unadjusted as adjusted coefficients between active and inactive adolescents are significant. Table 3 Multi-level linear regression results of continuous (logarithmically transformed) mean of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use, being bullied. * P < 0.001 Table 3 Multi-level linear regression results of continuous (logarithmically transformed) mean of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI    Total SDQ score  0.012  –0.001 to .025  0.152*  0.132 to 0.172  0.013  –0.000 to 0.026  0.111*  0.091 to 0.131  Emotional problems  0.090*  0.074 to 0.106  0.251*  0.228 to 0.274  0.073*  0.057 to 0.089  0.178*  0.155 to 0.201  Conduct problems  −0.047*  –0.060 to –0.033  0.032*  0.012 to 0.052  –0.032*  –0.045 to –0.019  0.025*  0.005 to 0.045  Hyperactivity  −0.016  –0.032 to 0.000  0.032**  0.012 to 0.052  –0.010  –0.026 to 0.006  0.028*  0.005 to 0.051  Peer problems  0.025*  0.009 to 0.041  0.203*  0.183 to 0.223  0.023*  0.007 to 0.039  0.152*  0.132 to 0.172  Pro-social behavior  –0.008*  –0.015 to –0.001  –0.056*  –0.066 to –0.046  –0.016  –0.023 to –0.009  −0.063*  −0.073 to −0.053  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use, being bullied. * P < 0.001 After the antilog procedure, the average total SDQ score of adolescents who were inactive was 1.12 times higher than that of adolescents who were active (β = 1.12; 95% CI: 1.10–1.14; P < 0.001; see table 4). In other words, inactive adolescents were likely to have psychosocial problems compared with active adolescents. Table 4 Multi-level linear regression results of continuous mean after the antilog procedure of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use and being bullied. * P < 0.001. Table 4 Multi-level linear regression results of continuous mean after the antilog procedure of the total SDQ score and the five subscales of the SDQ; unadjusted and adjusted models representing the effect of the difference on the SDQ score between active (reference), slightly active and inactive adolescents [β (95% (CI)]   Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95    Slightly active (ref. active)  Inactive (ref. active)  Slightly active (ref. active)  Inactive (ref. active)    Unadjusted model  Adjusted modela  SDQ-score  β  95% CI  β  95% CI  β  95% CI  β  95% CI  Total SDQ score  1.01  1.00 to 1.02  1.16*  1.14 to 1.19  1.01  1.00 to 1.03  1.12*  1.10 to 1.14  Emotional problems  1.09*  1.08 to 1.11  1.29*  1.26 to 1.32  1.08*  1.06 to 1.09  1.19*  1.17 to 1.22  Conduct problems  0.95*  0.94 to 0.97  1.03*  1.01 to 1.05  0.97*  0.96 to 0.98  1.03*  1.01 to 1.05  Hyperactivity  0.98  0.97 to 1.00  1.03*  1.01 to 1.05  0.99  0.97 to 1.01  1.03*  1.01 to 1.05  Peer problems  1.03*  1.01 to 1.04  1.23*  1.20 to 1.25  1.02*  1.01 to 1.04  1.16*  1.14 to 1.19  Pro-social behavior  0.99*  0.99 to 1.00  0.95*  0.94 to 0.96  0.98  0.98 to 0.99  0.94*  0.93 to 0.95  a Adjusted for gender, grade, household situation, ethnicity, urbanization, level of (secondary) education, use of soft drugs, hard drug use, recent use of alcohol, recent excessive alcohol use and being bullied. * P < 0.001. Additionally, it was relatively more common for inactive adolescents to have emotional problems (β = 1.19; 95% CI: 1.17–1.22; P < 0.001), conduct problems (β = 1.03; 95% CI: 1.01–1.05; P < 0.001), hyperactivity (β = 1.03; 95% CI: 1.00–1.05; P < 0.001), problems with peers (β = 1.16; 95% CI: 1.14–1.19; P < 0.001) and exhibit less pro-social behavior (β = 0.94; 95% CI: 0.93–0.95; P < 0.001) compared with active adolescents. The association between the average SDQ score of active adolescents and that of slightly active adolescents was almost statistically significant according to our threshold of 0.001 (β = 1.01; 95% CI: 1.00–1.03; P = 0.0012). Emotional problems (β = 1.08; 95% CI: 1.06–1.09; P < 0.001) were more common in slightly active adolescents compared with active adolescents. The weak but statistically significant association of conduct problems and problems with peers between these groups: 0.97 (95% CI: 0.96–0.98; P < 0.001) and 1.02 (95% CI: 1.01–1.04; P < 0.001), respectively. Discussion The aim of this study was to explore whether inactive and slightly active adolescents experience more psychosocial problems compared with active adolescents. Main findings of this study Our data provide evidence of an inverse association between the physical activity level of adolescents in the Netherlands and the extent of their psychosocial problems. Inactive adolescents experience more psychosocial problems than active adolescents. Similarly, inactive adolescents experience more emotional and conduct problems, hyperactivity, problems with peers and show less pro-social behavior. The differences between slightly active and active adolescents are less distinctive. For example, there is no statistical significant difference in total SDQ score between active and slightly active adolescents. Clinical relevance also plays a role in the interpretation of the observed association. Beforehand, we considered a result clinically relevant if the difference in adjusted regression (β) coefficient between active and inactive adolescents was greater than one point. Depending on the baseline value, an increase in one point on the SDQ scale could imply that an adolescent will be referred to a psychologist by the MPHS. The difference in weighted average total SDQ score between active and inactive adolescents is more than 1.5 point (table 2). This indicates, to our standards, a clinically relevant difference in psychosocial problems. However, due to the skewness of the distribution, we had to logarithmically transform the outcome measures. Therefore, the regression coefficients have to be interpreted as a ratio between those groups of adolescents instead of a difference. As a consequence, clinical relevance could not been deduced from these regression coefficients. What is already know on this topic The above-mentioned findings corroborates with earlier cross-sectional findings.11,13,14,16 These studies showed that lack of exercise was associated with various psychological problems. Psychosocial problems consist partly of psychological problems, making these findings relevant. However, these results are not in line with some longitudinal studies.24–26 Furthermore, Stavrakakis et al. focused on the onset of depressive disorders and physical activity but found no evidence of such an effect of physical activity on depression.22,38 This implies that the association we found in our data might be of limited value. On the other hand, the same Stavrakakis also studied the prospective relationship between physical activity and depressive symptoms in adolescents and found a weak statistically significant negative association.22,39 This shows that the relationship between physical activity and psychological health is complex. Genetic and biological factors, personality traits and environmental factors such as socialeconomic status and lifestyle habits interact and jointly affect psychosocial problems.22,38 In combination with these factors, sufficient physical exercise might prevent or mitigate psychosocial problems. What this study adds It was known in advance that most studies showed that physical exercise were associated with psychological problems. However, none of these studies focused on psychosocial problems. Psychosocial health is the outcome measure of the SDQ, widely used in youth care. This research has shown that physical exercise is also associated with less of psychosocial health. Finally, we used the data of the nationwide Dutch Youth Health Monitor conducted among almost 100 000 adolescents. Because of this number of participants, the results presented in this research are, as far as we know, unique. Limitations of this study The first weakness of this study is the cross-sectional design, which implies that a causal direction cannot be determined. Therefore, it may be that a lack of exercise causes psychosocial problems, but it is also possible that psychosocial problems lead to less exercise. Finally, it is also possible for an underlying factor to affect both physical exercises as psychosocial problems. The direction of causation can be established in a randomized controlled trial or in a prospective design. Second, the amount of physical activity was determined with a single-item question. This question covered the number of days per week of moderately intensive exercise for at least 1 h. Additional information about the exact amount and intensity was not collected. Information about the type of exercise, such as team sport or an individual activity, could deepen our understanding. Conclusion This study shows that physically active adolescents have fewer psychosocial problems than inactive adolescents. This association is not only statistically significant, but we think that it is also of clinical relevance. Inactive adolescents experience more emotional and conduct problems, hyperactivity, problems with peers and they show less pro-social behavior. Compared with active adolescents, slightly active adolescents experienced more emotional problems and problems with peers. However, slightly active adolescents experienced fewer conduct problems compared with active adolescents. Acknowledgements The data set of the Youth Health Monitor 2015 was provided by the Dutch National Institute for Public Health and Environment and 25 Municipal Public Health Service. Conflicts of interest: None declared. Key points Previous studies have demonstrated that 20% of the Dutch adolescents have some degree of psychosocial problems. Previous research investigated the relationship between physical activity and psychological health. However, none of these studies focused on psychosocial problems. We used the data of the nationwide Dutch Youth Health Monitor conducted among almost 100 000 adolescents. This study shows that physically active adolescents have fewer psychosocial problems than inactive adolescents. The amount of physical activity was determined with a single-item question. Information about the manner of exercise could deepen our understanding about the association between physical activity and psychosocial health. References 1 Theunissen MHC, Vogels AGC, Reijneveld SA. Work experience and style explain variation among pediatricians in the detection of children with psychosocial problems. Acad Pediatr  2012; 12: 495– 501. Google Scholar CrossRef Search ADS PubMed  2 Crone MR, Bekkema N, Wiefferink CH, et al.   Professional identification of psychosocial problems among children from ethnic minority groups: room for improvement. J Pediatr  2010; 156: 277– 84 e1. Google Scholar CrossRef Search ADS PubMed  3 Vannieuwenborg L, Buntinx F, De Lepeleire J. Presenting prevalence and management of psychosocial problems in primary care in Flanders. Arch Public Health  2015; 73: 10. Google Scholar CrossRef Search ADS PubMed  4 Anthony LG, Anthony BJ, Glanville DN, et al.   The relationships between parenting stress, parenting behaviour and preschoolers' social competence and behaviour problems in the classroom. Infant Child Dev  2005; 14: 133– 54. Google Scholar CrossRef Search ADS   5 Steinberg L, Lamborn SD, Dornbusch SM, et al.   Impact of parenting practices on adolescent achievement: authoritative parenting, school involvement, and encouragement to succeed. Child Dev  1992; 63: 1266– 81. Google Scholar CrossRef Search ADS PubMed  6 Rushton J, Bruckman D, Kelleher K. Primary care referral of children with psychosocial problems. Arch Pediatr Adolesc Med  2002; 156: 592– 8. Google Scholar CrossRef Search ADS PubMed  7 Smith JP, Smith GC. Long-term economic costs of psychological problems during childhood. Soc Sci Med  2010; 71: 110– 5. Google Scholar CrossRef Search ADS PubMed  8 Ten Have M, De Graaf R, Monshouwer K. [ Sports and Psychological Health. Results of the Netherlands Health Survey and Incidence Study (NEMESIS) ]. Utrecht: Trimbos Institute, 2009. 9 De Moor MH, Beem AL, Stubbe JH, et al.   Regular exercise, anxiety, depression and personality: a population-based study. Prev Med  2006; 42: 273– 9. Google Scholar CrossRef Search ADS PubMed  10 Goodwin RD. Association between physical activity and mental disorders among adults in the United States. Prev Med  2003; 36: 698– 703. Google Scholar CrossRef Search ADS PubMed  11 Kirkcaldy BD, Shephard RJ, Siefen RG. The relationship between physical activity and self-image and problem behaviour among adolescents. Soc Psych Psych Epidemol  2002; 37: 544– 50. Google Scholar CrossRef Search ADS   12 Muhsen K, Lipsitz J, Garty-Sandalon N, et al.   Correlates of generalized anxiety disorder: independent of co-morbidity with depression: findings from the first Israeli National Health Interview Survey (2003-2004). Soc Psych Psych Epidemiol  2008; 43: 898– 904. Google Scholar CrossRef Search ADS   13 Steptoe A, Wardle J, Fuller R, et al.   Leisure-time physical exercise: prevalence, attitudinal correlates, and behavioral correlates among young Europeans from 21 countries. Prev Med  1997; 26: 845– 54. Google Scholar CrossRef Search ADS PubMed  14 Tao FB, Xu ML, Kim SD, et al.   Physical activity might not be the protective factor for health risk behaviours and psychopathological symptoms in adolescents. J Paediatr Child Health  2007; 43: 762– 7. Google Scholar CrossRef Search ADS PubMed  15 Taylor MK, Pietrobon R, Pan D, et al.   Healthy People 2010 Physical activity guidelines and psychological symptoms: exercise from a large nationwide database. J Phys Act Health  2004; 1: 114– 30. Google Scholar CrossRef Search ADS   16 Ussher MH, Owen CG, Cook DG, et al.   The relationship between physical activity, sedentary behaviour and psychological wellbeing among adolescents. Soc Psychiatry Psychiatr Epidemiol  2007; 42: 851– 6. Google Scholar CrossRef Search ADS PubMed  17 Allison KR, Adlaf EM, Irving HM, et al.   Relationship of vigorous physical activity to psychologic distress among adolescents. J Adolescent Health  2005; 37: 164– 6. Google Scholar CrossRef Search ADS   18 Beard JR, Heathcote K, Brooks R, et al.   Predictors of mental disorders and their outcome in a community based cohort. Soc Psychiatry Psychiatr Epidemiol  2007; 42: 623– 30. Google Scholar CrossRef Search ADS PubMed  19 Brown WJ, Ford JH, Burton NW, et al.   Prospective study of physical activity and depressive symptoms in middle-aged women. Am J Prev Med  2005; 29: 265– 72. Google Scholar CrossRef Search ADS PubMed  20 Sanchez-Villegas A, Ara I, Guillen-Grima F, et al.   Physical activity, sedentary index, and mental disorders in the SUN cohort study. Med Sci Sports Exerc  2008; 40: 827– 34. Google Scholar CrossRef Search ADS PubMed  21 Strohle A, Hofler M, Pfister H, et al.   Physical activity and prevalence and incidence of mental disorders in adolescents and young adults. Psychol Med  2007; 37: 1657– 66. Google Scholar CrossRef Search ADS PubMed  22 McMahon EM, Corcoran P, O'Regan G, et al.   Physical activity in European adolescents and associations with anxiety, depression and well-being. Eur Child Adolesc Psychiatry  2016. 23 McPhie ML, Rawana JS. The effect of physical activity on depression in adolescence and emerging adulthood: a growth-curve analysis. J Adolesc  2015; 40: 83– 92. Google Scholar CrossRef Search ADS PubMed  24 De Moor MH, Boomsma DI, Stubbe JH, et al.   Testing causality in the association between regular exercise and symptoms of anxiety and depression. Arch Gen Psychiatry  2008; 65: 897– 905. Google Scholar CrossRef Search ADS PubMed  25 Kritz-Silverstein D, Barrett-Connor E, Corbeau C. Cross-sectional and prospective study of exercise and depressed mood in the elderly: the Rancho Bernardo study. Am J Epidemiol  2001; 153: 596– 603. Google Scholar CrossRef Search ADS PubMed  26 Wiles NJ, Jones GT, Haase AM, et al.   Physical activity and emotional problems amongst adolescents: a longitudinal study. Soc Psychiatry Psychiatr Epidemiol  2008; 43: 765– 72. Google Scholar CrossRef Search ADS PubMed  27 Goodman R. The Strengths and Difficulties Questionnaire: a research note. J Child Psychol Psychiatry  1997; 38: 581– 6. Google Scholar CrossRef Search ADS PubMed  28 Goodman A, Goodman R. Strengths and difficulties questionnaire as a dimensional measure of child mental health. J Am Acad Child Adolesc Psychiatry  2009; 48: 400– 3. Google Scholar CrossRef Search ADS PubMed  29 van Widenfelt BM, Goedhart AW, Treffers PD, et al.   Dutch version of the Strengths and Difficulties Questionnaire (SDQ). Eur Child Adolesc Psychiatry  2003; 12: 281– 9. Google Scholar CrossRef Search ADS PubMed  30 Muris P, Meesters C, van den Berg F. The Strengths and Difficulties Questionnaire (SDQ)–further evidence for its reliability and validity in a community sample of Dutch children and adolescents. Eur Child Adolesc Psychiatry  2003; 12: 1– 8. Google Scholar CrossRef Search ADS PubMed  31 Dutch Health Council. Guidelines for exercise 2017. 2017, Publication no. 2017/08. 32 Kemper HGC, Ooijendijk WTM, Stiggelbout M. Consensus on the Dutch Standard for Healthy Exercise. Ned Tijdschr Geneeskd  2000; 78: 180– 3. 33 Feng XQ, Wilson A. Do neighbourhood socioeconomic circumstances not matter for weight status among Australian men? Multilevel evidence from a household survey of 14 691 adults. BMJ Open  2015; 5: e007052. Google Scholar CrossRef Search ADS PubMed  34 Koezuka N, Koo M, Allison KR, et al.   The relationship between sedentary activities and physical inactivity among adolescents: results from the Canadian Community Health Survey. J Adolesc Health  2006; 39: 515– 22. Google Scholar CrossRef Search ADS PubMed  35 Chang BH, Pocock S. Analyzing data with clumping at zero. An example demonstration. J Clin Epidemiol  2000; 53: 1036– 43. Google Scholar CrossRef Search ADS PubMed  36 Lachin JM, McGee PL, Greenbaum CJ, et al.   Sample size requirements for studies of treatment effects on beta-cell function in newly diagnosed type 1 diabetes. PLoS One  2011; 6: e26471. Google Scholar CrossRef Search ADS PubMed  37 Schlomer BJ, Copp HL. Secondary data analysis of large data sets in urology: successes and errors to avoid. J Urol  2014; 191: 587– 96. Google Scholar CrossRef Search ADS PubMed  38 Stavrakakis N, Roest AM, Verhulst F, et al.   Physical activity and onset of depression in adolescents: a prospective study in the general population cohort TRAILS. J Psychiatr Res  2013; 47: 1304– 8. Google Scholar CrossRef Search ADS PubMed  39 Stavrakakis N, de Jonge P, Ormel J, et al.   Bidirectional prospective associations between physical activity and depressive symptoms. The TRAILS Study. J Adolesc Health  2012; 50: 503– 8. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The European Journal of Public HealthOxford University Press

Published: Jan 8, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off