International trends in ‘bottom-end’ inequality in adolescent physical activity and nutrition: HBSC study 2002–2014

International trends in ‘bottom-end’ inequality in adolescent physical activity and... Abstract Background In spite of many positive trends that have emerged in the health of young people, adolescents from more affluent groups continue to experience more favourable health outcomes. There are no groups that are more vulnerable than those who report very poor (‘bottom-end’) indicators of health behaviour. The present study investigated the role of socio-economic factors as potential determinants of bottom-end health behaviours pertaining to physical activity and diet. Methods Our analysis incorporated health data for some 700 000 15-year-old adolescents in 34 countries. The data source was four cycles of the Health Behaviour in School-aged Children (HBSC) study (2001/2002, 2005/2006, 2009/2010 and 2013/2014). As per UNICEF precedents, adolescents whose health behaviour scores were below the mean of the lower half of the distribution fell into the ‘bottom-end’ on this indicator. Results Adolescents from less affluent families were much more likely to report being in the bottom-end of the distribution of these health indicators. Large, persistent and widespread socio-economic gradients existed for physical activity and healthy eating, while the findings were mixed for unhealthy eating. Such socio-economic inequalities were largely stable or widened for physical activity and healthy eating, while inequalities in unhealthy eating narrowed. Conclusion Although it is important to continue monitoring average levels of adolescent health, national and international policies need to pay attention to the concentration of poor health outcomes among adolescents from less affluent families and to redress social inequalities in adolescent health behaviour. Introduction Contemporary cross-national studies demonstrate that, in most countries, the majority of young people report a very good health status. This is reflected by indicators of general health,1 healthy dietary practices,2 engagement in physical activity3 and abstinence from substance misuse.4–6 Yet, inequalities in health persist. These are especially disadvantageous to adolescents with a lower socio-economic status (SES).7 Socio-economic gradients in adolescent health have persisted over recent decades.4,8,9 Among adolescents, such inequalities have been identified for self-rated health; psychological health; academic performance;10 mental health problems;11 life satisfaction;12overweight and obesity; and its behavioural determinant of physical inactivity.13,14 There is evidence of an increasing socio-economic gap in such health indicators during the last decade,8,9 although the available research base is still scant. Consideration of these inequalities and their impacts remain a high priority for health policy.15 Relatively few studies have focused on international trends in adolescent health inequalities.8,9 Few cross-national studies have investigated the experiences of adolescents who report substantially worse health outcomes (the so-called ‘bottom-end’ of health16–18) relative to others. Focus on adolescents at risk for poor health is important as such inequalities tend to perpetuate into adulthood.15 While economic arguments suggest that investing in the early childhood years is an efficient strategy to build a productive future workforce,19 similar arguments are seldom heard with respect to adolescence, although this remains a crucial developmental period for a healthy transition to adulthood. There is also evidence that interventions that are implemented during adolescence can mitigate adverse effects of poor well-being during earlier childhood.20 In a large cross-national study that spanned 12 years (2001/2002–2013/2014) and involved some 700 000 adolescents from 31 European and North American countries, we studied the extent to which their SES, relative to their peers in the same country, was associated with being in the bottom-end of different indicators of health behaviours. Here, we focused on indicators of physical activity and diet. Physical inactivity and the absence of a balanced diet are key reasons for increases in childhood obesity21,22 and other indicators of short-term and chronic morbidity.23–30 Our primary objective was to estimate the magnitude of the association between family SES and poor (‘bottom-end’) adolescent physical activity and dietary behaviours across 34 countries. Moreover, if there was evidence of such an association, we aimed to determine whether the magnitude of this socio-economic gradient changed over time in the different countries (2002–2014). This foundational information would inform social policies and health promotion strategies, internationally. Methods ‘Health Behaviour in School-aged Children’ (HBSC) is a cross-sectional school-based study that has been conducted every 4 year since 1982. HBSC aims to increase the understanding of adolescent health and well-being as well as the health behaviours and social environments that contribute to such outcomes.5 Data are collected from 11- to 15-year-old school children according to a common international protocol. Each member country obtained ethics clearance to conduct the survey from a university-based review board or equivalent body. A detailed description of the aims and theoretical framework of the HBSC study can be found elsewhere.4 The current analysis used data from the last four HBSC cycles in 2001/2002, 2005/2006, 2009/2010 and 2013/2014 for countries of the European Union and/or the Organization for Economic Co-operation and Development. After excluding countries where 10% or more of the students had missing information on key indicators, we report findings from 700 000 young people in 34 countries, with 26 countries present in all four cycles. The following countries had one or more cycles missing for at least one indicator: Belgium (2001/2002), Bulgaria (2001/2002 and 2009/2010), Germany (2001/2002), Iceland (2001/2002), Israel (2013/2014), Luxembourg (2001/2002), Malta (2005/2006 and 2009/2010), Romania (2001/2002), Slovakia (2001/2002 and 2005/2006), Switzerland (2001/2006), Turkey (2001/2002 and 2013/2014) and the United States (2013/2014). Socio-economic status The Family Affluence Scale (FAS) is a measure of material family wealth developed as an indicator of absolute level of SES.23,24 Between 2001/2002 and 2009/2010, the HBSC FAS scale was comprised of reports for the following items: ownership of a family car (0, 1, 2 or more); own bedroom (no = 0, yes = 1); family holidays during the past 12 months (0, 1, 2, 3 or more); and family computer(s) (0, 1, 2, 3 or more). These items are combined to produce a composite score ranging from 0 (low affluence) to 9 (high affluence). In 2013/2014, two more items were introduced to this scale, numbers of bathrooms and ownership of a dishwasher.24 As per precedent,31 cycle-specific measures of FAS were transformed into a continuous proportional rank score ranging from 0 to 1, separately by country, with the country sample means set at 0.5. These country-specific ridit scores reflect the proportion of respondents with lower family affluence, with higher values reflecting higher levels of SES relative to others within the country. In regression models of health, one unit on the ridit scale refers to the difference between the least affluent and most affluent adolescents in the country.8 Physical activity Moderate-to-vigorous physical activity (MVPA) was measured by asking on how many days over the past seven participants were physically active for at least 60 min. The survey item defined MVPA as ‘any activity that increases your heart rate and makes you get out of breath some of the time’ and includes examples such as running and brisk walking.24 Response options ranged from 0 days to 7 days. The item has been used in the HBSC since 2001/2002 and correlates highly with a general question about physical activities.3 The measure reflects current policy recommendations for children’s physical activity.25 Healthy diet Following precedent,26 two items on fruit and vegetable consumption from an abbreviated food frequency questionnaire were combined into a healthy eating index. Each item was first re-coded from an ordinal to a ratio scale as follows: ‘never’= 0, ‘less than once a week’ = 0.25, ‘once a week’ = 1, ‘2–4 days a week’ = 3, ‘5–6 days a week’ = 5.5 and ‘once a day, every day’ and ‘more than once a day, every day’ = 7. The two items were then added together so that the index ranged from 0 to 14, with 0 corresponding to never eating fruit or vegetables and 14 to eating both fruit and vegetables at least once a day. Unhealthy diet Items on sweets and sugar-added soft drinks consumption similarly contributed to an unhealthy eating index. Items were first reverse-coded as follows and then summed together: ‘never’ = 7, ‘less than once a week’ = 5.5, ‘once a week’ = 3, ‘2–4 times/week’ = 1, ‘5–6 days a week’ = 0.25 and ‘once a day, every day’ and ‘more than once a day, every day’ = 0. In the composite 0–14 scale, 0 denotes consuming both sweets and sugared drinks at least once a day and 14 refers to never consuming sweets or sugary drinks. Indicators of ‘bottom-end’ health To identify adolescents who have substantially poorer health behaviours relative to their peers, we estimated the mean in the lower half of the distribution for each indicator separately, by country and year. Adolescents who fell below this threshold reported worse outcomes than an average respondent among the least well performing half of their population. This method has been applied in past HBSC,16 UNICEF17,18 and other international reports.27 The present study focuses on the individual determinants of scoring in the bottom-end group (separately by indicator, country and year) rather than on the country-level dispersion of health scores in the lower half of the distribution [see 27]. Analysis Our analysis focused on the strength of associations between the country-specific ridit-transformed FAS score and bottom-end health outcomes. We used linear probability models (i.e. ordinary least squares regressions with binary dependent variables) separately for each indicator, country and survey year. Such models are computationally straightforward and easy to interpret in terms of differences in probabilities of the outcome, but since they can produce biased estimates28 we also replicated all analyses using logistic regression. The findings were qualitatively identical (see the online appendix). All models controlled for age and gender and were fitted separately by country, using sample weights. Standard errors were adjusted to reflect the complex sampling structure of the surveys, with schools identified as primary sampling units and regions as strata (where relevant, i.e. in Belgium and the United Kingdom). We report on the effects of a one-unit difference in the ridit score for family affluence (the difference between the least and most affluent adolescents) on the probability of being in the bottom-end group for each indicator, controlling for age and gender. To investigate whether the effects of family affluence have changed significantly (P < 0.05) over the last four survey cycles, we included a trend variable (with 2001/2002 cycle as referent) with interaction terms between this trend variable and each of the three other predictors (i.e. family affluence, age and gender). Results There was substantial cross-national variation in the size of the bottom-end group of adolescents (i.e. those scoring below the mean of the lower half of the distribution in their country) to define the bottom-end group. The share of adolescents that ‘fell behind’ on MVPA ranged from 10% in Portugal to 27% in Belgium, while the proportion of those that ‘fell behind’ on healthy eating (fruit and vegetable consumption) varied from 16% in Belgium and Switzerland to 26% in Hungary and Spain. While one in four adolescents (25%) scored poorly on unhealthy eating (abstinence from sweets and sugary drinks) in the combined sample, this varied from 12% in Sweden to 33% in Norway. Table 1 shows large and widespread socio-economic inequalities in physical activity and healthy eating, with mixed results for unhealthy eating. In all 31 countries in the 2013/2014 analysis (note that there are 34 countries overall, but only 31 in 2013/2014), adolescents from more affluent households were significantly (P < 0.05) less likely to fall behind in MVPA, with the FAS gradient ranging from −8.3% in Finland (best performing) to −21.9% in Luxembourg (worst performing). Thus, the probability of scoring in the bottom-end group in MVPA is 21.9% points lower for adolescents from the most affluent families in Luxembourg than for their peers from the least affluent families. Similarly, adolescents in the lowest SES strata were more likely to fall further behind in healthy eating in all countries except Malta and Romania, where there were no statistically significant differences identified. The United Kingdom showed the largest negative gradient in healthy eating: the poorest adolescents were 21.3% points more likely to fall furthest behind. Hungary posted the largest negative gradient in unhealthy eating: adolescents from the most affluent families were 16.1% points less likely to consume excess sugar. In contrast, the most affluent adolescents in Estonia were 7.1% points more likely to engage in unhealthy eating. Table 1 Effect of family affluence on the probability of falling in the bottom-end group, controlling for age and gender (HBSC 2013/2014) Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Notes: Percentage point difference between the most and the least affluent adolescents; Standard errors adjusted for clustering at the school level and stratification by region (Belgium and the United Kingdom); Statistically significant effects (P < 0.05) in bold. Table 1 Effect of family affluence on the probability of falling in the bottom-end group, controlling for age and gender (HBSC 2013/2014) Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Notes: Percentage point difference between the most and the least affluent adolescents; Standard errors adjusted for clustering at the school level and stratification by region (Belgium and the United Kingdom); Statistically significant effects (P < 0.05) in bold. In the vast majority of the countries studied, the sizeable and statistically significant socio-economic gradient in MVPA remained stable over time (figure 1). In six countries, it became significantly (P < 0.05) larger in absolute terms between 2001/2002 and 2013/2014: Belgium, Italy, Latvia, the Netherlands, Sweden and the United Kingdom. Figure 1 View largeDownload slide FAS gradient in bottom-end physical activity between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Figure 1 View largeDownload slide FAS gradient in bottom-end physical activity between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 The FAS gradient in healthy eating remained stable over time in most of the countries studied (figure 2). In Canada and the United Kingdom this gradient increased, suggesting that adolescents from lower socio-economic backgrounds were increasingly less likely to consume fruit and vegetables when compared with their peers. In three other countries—Latvia, Lithuania and Romania—the FAS gradient has decreased. Although in Latvia and Lithuania adolescents from lower-affluence families were still significantly (P < 0.05) less likely to eat healthily, in Romania socio-economic inequalities were no longer statistically significant in 2013/2014. Figure 2 View largeDownload slide FAS gradient in bottom-end healthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Figure 2 View largeDownload slide FAS gradient in bottom-end healthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Conversely, in the majority of the countries analysed, there were no statistically significant socio-economic inequalities in unhealthy eating in 2013/2014. In eight countries, adolescents from more affluent households were less likely to report frequent consumption of sweets and sugary drinks, while in three others (Estonia, Latvia and the Netherlands), they were more likely to do so. While in many countries, there was not a significant (P < 0.05) FAS gradient in unhealthy eating between 2001/2002 and 2013/2014 (figure 3), higher consumption of sugar in less affluent households persisted over time in France, Ireland and the United Kingdom. Figure 3 View largeDownload slide FAS gradient in bottom-end unhealthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Figure 3 View largeDownload slide FAS gradient in bottom-end unhealthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 The FAS gradient in unhealthy eating changed significantly (P < 0.05) between 2001/2002 and 2013/2014 in eight countries. In six of these—Estonia, Latvia, Lithuania, Poland, Portugal and Romania—adolescents from more affluent backgrounds were more likely to report a higher frequency of unhealthy eating than their peers in 2001/2002, but these differences either narrowed or disappeared by 2013/2014. In the remaining two countries, Belgium and Hungary, adolescents from less affluent backgrounds were increasingly more likely to report a higher frequency of unhealthy eating. Discussion Adolescence is a critical period of transition in the life course, characterized by biological, psychological and relational changes. Such transitions are of fundamental importance to health15 and set the stage for future patterns of adult inequalities.29 In the vast majority of studied countries between 2011/2012 and 2013/2014, adolescents from relatively low SES families had a greater likelihood of falling furthest behind in health, particularly with respect to reported physical activity and healthy eating. For unhealthy eating another pattern emerged, with no indication of an association with SES for the majority of the countries and mixed results for the rest. Sweets and soft drink consumption may be associated with knowledge about healthy diets, which is typically higher in more affluent socio-economic groups.30 Indeed, in eight countries (i.e. Belgium, Denmark, France, Hungary, Iceland, Ireland, Spain and the United Kingdom), young people from low socio-economic groups were overrepresented in the bottom-end of unhealthy eating in 2013/2014. In four of these (i.e. Belgium, France, Spain and the United Kingdom), the pattern has persisted since 2001/2002. Meanwhile, in another six countries (i.e. Croatia, Estonia, Latvia, Lithuania, Poland and Portugal) adolescents from more affluent backgrounds were more likely to report a higher frequency of unhealthy eating than their peers in 2001/2002, although the association decreased or disappeared by 2013/2014. This is consistent with historical studies that documented a positive association between family affluence and frequent soft drink consumption among adolescents in Central and Eastern Europe (CEE) at the turn of the century, possibly as an indicator of the ability to afford such luxuries and hence relative wealth.32 Inequalities in adolescent health were widening in some of the countries studied. Socio-economic inequalities in MVPA increased in six countries during the 12-year study period (Belgium, Italy, Latvia, the Netherlands, Sweden and the United Kingdom). In Canada and the United Kingdom, the association between SES and adolescent healthy eating became more pronounced over time, as did unhealthy eating in Belgium and Hungary. Widening national income inequalities tend to increase socio-economic differences in health, perpetuating the socio-economic divide.8 Conversely, in some (mostly CEE) countries, inequalities in adolescent health decreased considerably between 2001/2002 and 2013/2014. Adolescents in Latvia and Lithuania from less affluent families were less likely to eat healthily, but this association weakened significantly (P < 0.05) between 2001/2002 and 2013/2014. For Romania, socio-economic inequalities in healthy eating disappeared in the same period. These results suggest that social gradients in health can evolve over time. This study has several potential limitations. First, the cross-sectional design of the HBSC study precludes the establishment of the temporal sequence of events, limiting claims on causality. It is unlikely, however, that the outcomes of physical activity and diet would precede (or be determinants) of SES. Second, the indicators of adolescent health used here, although standardized and validated for cross-national comparison,26,33 come with their own inherent limitations. The indicators of fruit, vegetables, sweets and sugary drink consumption are all based on frequency of intake, rather than amounts consumed.2 Thus, we do not know whether, e.g. adolescents who reported daily consumption of fruit and vegetables actually consumed five portions of fruit and vegetables a day, but our measures are the best proxies for healthy and unhealthy eating afforded by our data. Our measure of SES relies on an assumption that ‘the scores of the [FAS] scale can be used to rank individuals and groups along a latent continuum of material wealth’.31 Although FAS has been validated against other measures of family SES, such as parental occupation,23 the scale is subject to measurement error; and it is very challenging to assess this construct by self-report in populations as young as 11 years. Finally, not all countries were present in each of the four survey cycles, limiting the cross-country comparability of the results. In conclusion, this study establishes that socio-economic inequalities in adolescent physical activity and dietary behaviours are large and stable across countries and over time. These are important findings because they suggest two common mechanisms—physical activity and diet—by which health inequalities emerge and adolescents become socially disadvantaged within and across different cultures and societies. The results indicate a fundamental unfairness16–18 that affects the most socio-economically disadvantaged adolescents in almost all societies; an unfairness that is consistent and persistent and sets the stage for negative health trajectories. These findings point to a universal need for public health interventions focused on physical activity and diet, including social policies that specifically target these and other aspects of health in our most disadvantaged children, as priorities, nationally and internationally. Acknowledgements The Health Behaviour in School-aged Children (HBSC) study is a World Health Organization collaborative study and is supported by each member country of the HBSC network (www.hbsc.org). The HBSC study is coordinated internationally by Dr Joanna Inchley, University of St. Andrews, Scotland, with international data coordination performed by Dr Oddrun Samdal, University of Bergen, Norway. Funding No external funding. Conflicts of interest: None declared. Key points Few cross-national studies have investigated the experiences of adolescents who report substantially worse health outcomes relative to others. This study reveals large, widespread and persistent socio-economic inequalities in the risks of the poorest adolescent health behaviours with respect to reported physical activity and healthy eating both across 34 industrialized countries and over time (between 2002 and 2014). Inequalities in unhealthy eating are less prevalent and may change their direction over time. References 1 Cavallo F, Dalmasso P, Ottová-Jordan V, et al.   Trends in self-rated health in European and North-American adolescents from 2002 to 2010 in 32 countries. Eur J Public Health  2015; 25: 13– 5. 2 Vereecken C, Pedersen TP, Ojala K, et al.   Fruit and vegetable consumption trends among adolescents from 2002 to 2010 in 33 countries. Eur J Public Health  2015; 25: 16– 9. http://dx.doi.org/10.1093/eurpub/ckv012 Google Scholar CrossRef Search ADS PubMed  3 Kalman M, Inchley J, Sigmundova D, et al.   Secular trends in moderate-to-vigorous physical activity in 32 countries from 2002 to 2010: a cross-national perspective. Eur J Public Health  2015; 25(suppl 2): 37– 40. Google Scholar CrossRef Search ADS PubMed  4 Currie C, Zanotti C, Morgan A, et al.   Social Determinants of Health and Well-Being Among Young People . Copenhagen: World Health Organization Regional Office for Europe, 2012. 5 de Looze M, Raaijmakers Q, Ter Bogt T, et al.   Decreases in adolescent weekly alcohol use in Europe and North America: evidence from 28 countries from 2002 to 2010. Eur J Public Health  2015; 25: 69– 72. http://dx.doi.org/10.1093/eurpub/ckv031 6 Hublet A, Bendtsen P, de Looze ME, et al.   Trends in the co-occurrence of tobacco and cannabis use in 15-year-olds from 2002 to 2010 in 28 countries of Europe and North America. Eur J Public Health  2015; 25: 73– 5. http://dx.doi.org/10.1093/eurpub/ckv032 7 Kuntsche E, Ravens-Sieberer U. Monitoring adolescent health behaviours and social determinants cross-nationally over more than a decade: introducing the Health Behaviour in School-aged Children (HBSC) study supplement on trends. Eur J Public Health  2015; 25: 1– 3. http://dx.doi.org/10.1093/eurpub/ckv009 Google Scholar CrossRef Search ADS PubMed  8 Elgar FJ, Pförtner T-K, Moor I, et al.   Socioeconomic inequalities in adolescent health 2002–2010: a time-series analysis of 34 countries participating in the Health Behaviour in School-aged Children study. The Lancet  2015; 385: 2088– 95. Google Scholar CrossRef Search ADS   9 Moor I, Rathmann K, Lenzi M, et al.   Socioeconomic inequalities in adolescent smoking across 35 countries: a multilevel analysis of the role of family, school and peers. Eur J Public Health  2015; cku244. 10 Magklara K, Skapinakis P, Niakas D, et al.   Socioeconomic inequalities in general and psychological health among adolescents: a cross-sectional study in senior high schools in Greece. Int J Equity Health  2010; 9: 3. http://dx.doi.org/10.1186/1475-9276-9-3 Google Scholar CrossRef Search ADS PubMed  11 Reiss F. Socioeconomic inequalities and mental health problems in children and adolescents: a systematic review. Soc Sci Med  2013; 90: 24– 31. http://dx.doi.org/10.1016/j.socscimed.2013.04.026 Google Scholar CrossRef Search ADS PubMed  12 Moor I, Lampert T, Rathmann K, et al.   Explaining educational inequalities in adolescent life satisfaction: do health behaviour and gender matter? Int J Public Health  2014; 59: 309– 17. Google Scholar CrossRef Search ADS PubMed  13 Frederick CB, Snellman K, Putnam RD. Increasing socioeconomic disparities in adolescent obesity. Proc Natl Acad Sci USA  2014; 111: 1338– 42. http://dx.doi.org/10.1073/pnas.1321355110 Google Scholar CrossRef Search ADS PubMed  14 Stalsberg R, Pedersen AV. Effects of socioeconomic status on the physical activity in adolescents: a systematic review of the evidence. Scand J Med Sci Sports  2010; 20: 368– 83. http://dx.doi.org/10.1111/j.1600-0838.2009.01047.x Google Scholar CrossRef Search ADS PubMed  15 Viner RM, Ozer EM, Denny S, et al.   Adolescence and the social determinants of health. The Lancet  2012; 379: 1641– 52. http://dx.doi.org/10.1016/S0140-6736(12)60149-4 Google Scholar CrossRef Search ADS   16 Currie C, Menchini L, Richardson D, et al.   Comparing inequality in the well-being of children in economically advanced countries: a methodology. Innocenti Work Paper:IWP-2010-19, 2011. 17 UNICEF. The Children Left behind: A League Table of Inequality in Child Well-Being in the World’s Rich Countries . Florence: UNICEF Innocenti Research Centre, 2010. 18 UNICEF. Fairness for Children: A league table of inequality in child well-being in rich countries, Innocenti Report Card 13. Florence: UNICEF Office of Research, 2016. 19 Heckman JJ. Skill formation and the economics of investing in disadvantaged children. Science  2006; 312: 1900– 2. http://dx.doi.org/10.1126/science.1128898 Google Scholar CrossRef Search ADS PubMed  20 McDaid D, Park A-L, Currie C, Zanotti C. Investing in the wellbeing of young people: making the economic case. In: McDaid D, Cooper C, editors. Economics of Wellbeing . Wellbeing: A Complete Reference Guide (Vol. 5). Oxford: Wiley-Blackwell, 2014, 181– 214. 21 Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: public-health crisis, common sense cure. The Lancet  2002; 360: 473– 82. http://dx.doi.org/10.1016/S0140-6736(02)09678-2 Google Scholar CrossRef Search ADS   22 Nishtar S, Gluckman P, Armstrong T. Ending childhood obesity: a time for action. The Lancet  2016; 387: 825– 7. http://dx.doi.org/10.1016/S0140-6736(16)00140-9 Google Scholar CrossRef Search ADS   23 Currie C, Molcho M, Boyce W, et al.   Researching health inequalities in adolescents: the development of the Health Behaviour in School-Aged Children (HBSC) family affluence scale. Soc Sci Med  2008; 66: 1429– 36. http://dx.doi.org/10.1016/j.socscimed.2007.11.024 Google Scholar CrossRef Search ADS PubMed  24 Currie C, Inchley J, Molcho M, et al.   Health Behaviour in School-Aged Children (HBSC) Study Protocol: Background, Methodology and Mandatory Items for the 2013/14 Survey . St Andrews: CAHRU, 2014. 25 WHO. Information sheet: Global recommendations on physical activity for health 5-17 years old [Internet]. World Health Organisation, 2011. Available at: http://www.who.int/dietphysicalactivity/publications/recommendations5_17years/en/ (11 March 2016, date last accessed). 26 Vereecken C, Rossi S, Giacchi MV, Maes L. Comparison of a short food-frequency questionnaire and derived indices with a seven-day diet record in Belgian and Italian children. Int J Public Health  2008; 53: 297– 305. Google Scholar PubMed  27 Chzhen Y, Bruckauf Z, Ng K, et al.   Inequalities in adolescent health and life satisfaction: Evidence form the Health Behaviour in School-Aged Children study. Florence: UNICEF Office of Research, 2016. Report no: 2016–11. 28 Horrace WC, Oaxaca RL. Results on the bias and inconsistency of ordinary least squares for the linear probability model. Econ Lett  2006; 90: 321– 7. http://dx.doi.org/10.1016/j.econlet.2005.08.024 Google Scholar CrossRef Search ADS   29 Patton GC, Coffey C, Cappa C, et al.   Health of the world’s adolescents: a synthesis of internationally comparable data. The Lancet  2012; 379: 1665– 75. Google Scholar CrossRef Search ADS   30 Hulshof K, Brussaard JH, Kruizinga AG, et al.   Socio-economic status, dietary intake and 10 y trends: the Dutch National Food Consumption Survey. Eur J Clin Nutr  2003; 57: 128– 37. http://dx.doi.org/10.1038/sj.ejcn.1601503 Google Scholar CrossRef Search ADS PubMed  31 Torsheim T, Currie C, Boyce W, et al.   Material deprivation and self-rated health: a multilevel study of adolescents from 22 European and North American countries. Soc Sci Med  2004; 59: 1– 12. http://dx.doi.org/10.1016/j.socscimed.2003.09.032 Google Scholar CrossRef Search ADS PubMed  32 Vereecken C, Inchley J, Subramanian SV, et al.   The relative influence of individual and contextual socio-economic status on consumption of fruit and soft drinks among adolescents in Europe. Eur. J Public Health  2005; 15: 224– 32. 33 Prochaska JJ, Sallis JF, Long B. A physical activity screening measure for use with adolescents in primary care. Arch Pediatr Adolesc Med  2001; 155: 554– 9. http://dx.doi.org/10.1001/archpedi.155.5.554 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. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The European Journal of Public Health Oxford University Press

International trends in ‘bottom-end’ inequality in adolescent physical activity and nutrition: HBSC study 2002–2014

Loading next page...
 
/lp/ou_press/international-trends-in-bottom-end-inequality-in-adolescent-physical-CjYVW0rfX4
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/ckx237
Publisher site
See Article on Publisher Site

Abstract

Abstract Background In spite of many positive trends that have emerged in the health of young people, adolescents from more affluent groups continue to experience more favourable health outcomes. There are no groups that are more vulnerable than those who report very poor (‘bottom-end’) indicators of health behaviour. The present study investigated the role of socio-economic factors as potential determinants of bottom-end health behaviours pertaining to physical activity and diet. Methods Our analysis incorporated health data for some 700 000 15-year-old adolescents in 34 countries. The data source was four cycles of the Health Behaviour in School-aged Children (HBSC) study (2001/2002, 2005/2006, 2009/2010 and 2013/2014). As per UNICEF precedents, adolescents whose health behaviour scores were below the mean of the lower half of the distribution fell into the ‘bottom-end’ on this indicator. Results Adolescents from less affluent families were much more likely to report being in the bottom-end of the distribution of these health indicators. Large, persistent and widespread socio-economic gradients existed for physical activity and healthy eating, while the findings were mixed for unhealthy eating. Such socio-economic inequalities were largely stable or widened for physical activity and healthy eating, while inequalities in unhealthy eating narrowed. Conclusion Although it is important to continue monitoring average levels of adolescent health, national and international policies need to pay attention to the concentration of poor health outcomes among adolescents from less affluent families and to redress social inequalities in adolescent health behaviour. Introduction Contemporary cross-national studies demonstrate that, in most countries, the majority of young people report a very good health status. This is reflected by indicators of general health,1 healthy dietary practices,2 engagement in physical activity3 and abstinence from substance misuse.4–6 Yet, inequalities in health persist. These are especially disadvantageous to adolescents with a lower socio-economic status (SES).7 Socio-economic gradients in adolescent health have persisted over recent decades.4,8,9 Among adolescents, such inequalities have been identified for self-rated health; psychological health; academic performance;10 mental health problems;11 life satisfaction;12overweight and obesity; and its behavioural determinant of physical inactivity.13,14 There is evidence of an increasing socio-economic gap in such health indicators during the last decade,8,9 although the available research base is still scant. Consideration of these inequalities and their impacts remain a high priority for health policy.15 Relatively few studies have focused on international trends in adolescent health inequalities.8,9 Few cross-national studies have investigated the experiences of adolescents who report substantially worse health outcomes (the so-called ‘bottom-end’ of health16–18) relative to others. Focus on adolescents at risk for poor health is important as such inequalities tend to perpetuate into adulthood.15 While economic arguments suggest that investing in the early childhood years is an efficient strategy to build a productive future workforce,19 similar arguments are seldom heard with respect to adolescence, although this remains a crucial developmental period for a healthy transition to adulthood. There is also evidence that interventions that are implemented during adolescence can mitigate adverse effects of poor well-being during earlier childhood.20 In a large cross-national study that spanned 12 years (2001/2002–2013/2014) and involved some 700 000 adolescents from 31 European and North American countries, we studied the extent to which their SES, relative to their peers in the same country, was associated with being in the bottom-end of different indicators of health behaviours. Here, we focused on indicators of physical activity and diet. Physical inactivity and the absence of a balanced diet are key reasons for increases in childhood obesity21,22 and other indicators of short-term and chronic morbidity.23–30 Our primary objective was to estimate the magnitude of the association between family SES and poor (‘bottom-end’) adolescent physical activity and dietary behaviours across 34 countries. Moreover, if there was evidence of such an association, we aimed to determine whether the magnitude of this socio-economic gradient changed over time in the different countries (2002–2014). This foundational information would inform social policies and health promotion strategies, internationally. Methods ‘Health Behaviour in School-aged Children’ (HBSC) is a cross-sectional school-based study that has been conducted every 4 year since 1982. HBSC aims to increase the understanding of adolescent health and well-being as well as the health behaviours and social environments that contribute to such outcomes.5 Data are collected from 11- to 15-year-old school children according to a common international protocol. Each member country obtained ethics clearance to conduct the survey from a university-based review board or equivalent body. A detailed description of the aims and theoretical framework of the HBSC study can be found elsewhere.4 The current analysis used data from the last four HBSC cycles in 2001/2002, 2005/2006, 2009/2010 and 2013/2014 for countries of the European Union and/or the Organization for Economic Co-operation and Development. After excluding countries where 10% or more of the students had missing information on key indicators, we report findings from 700 000 young people in 34 countries, with 26 countries present in all four cycles. The following countries had one or more cycles missing for at least one indicator: Belgium (2001/2002), Bulgaria (2001/2002 and 2009/2010), Germany (2001/2002), Iceland (2001/2002), Israel (2013/2014), Luxembourg (2001/2002), Malta (2005/2006 and 2009/2010), Romania (2001/2002), Slovakia (2001/2002 and 2005/2006), Switzerland (2001/2006), Turkey (2001/2002 and 2013/2014) and the United States (2013/2014). Socio-economic status The Family Affluence Scale (FAS) is a measure of material family wealth developed as an indicator of absolute level of SES.23,24 Between 2001/2002 and 2009/2010, the HBSC FAS scale was comprised of reports for the following items: ownership of a family car (0, 1, 2 or more); own bedroom (no = 0, yes = 1); family holidays during the past 12 months (0, 1, 2, 3 or more); and family computer(s) (0, 1, 2, 3 or more). These items are combined to produce a composite score ranging from 0 (low affluence) to 9 (high affluence). In 2013/2014, two more items were introduced to this scale, numbers of bathrooms and ownership of a dishwasher.24 As per precedent,31 cycle-specific measures of FAS were transformed into a continuous proportional rank score ranging from 0 to 1, separately by country, with the country sample means set at 0.5. These country-specific ridit scores reflect the proportion of respondents with lower family affluence, with higher values reflecting higher levels of SES relative to others within the country. In regression models of health, one unit on the ridit scale refers to the difference between the least affluent and most affluent adolescents in the country.8 Physical activity Moderate-to-vigorous physical activity (MVPA) was measured by asking on how many days over the past seven participants were physically active for at least 60 min. The survey item defined MVPA as ‘any activity that increases your heart rate and makes you get out of breath some of the time’ and includes examples such as running and brisk walking.24 Response options ranged from 0 days to 7 days. The item has been used in the HBSC since 2001/2002 and correlates highly with a general question about physical activities.3 The measure reflects current policy recommendations for children’s physical activity.25 Healthy diet Following precedent,26 two items on fruit and vegetable consumption from an abbreviated food frequency questionnaire were combined into a healthy eating index. Each item was first re-coded from an ordinal to a ratio scale as follows: ‘never’= 0, ‘less than once a week’ = 0.25, ‘once a week’ = 1, ‘2–4 days a week’ = 3, ‘5–6 days a week’ = 5.5 and ‘once a day, every day’ and ‘more than once a day, every day’ = 7. The two items were then added together so that the index ranged from 0 to 14, with 0 corresponding to never eating fruit or vegetables and 14 to eating both fruit and vegetables at least once a day. Unhealthy diet Items on sweets and sugar-added soft drinks consumption similarly contributed to an unhealthy eating index. Items were first reverse-coded as follows and then summed together: ‘never’ = 7, ‘less than once a week’ = 5.5, ‘once a week’ = 3, ‘2–4 times/week’ = 1, ‘5–6 days a week’ = 0.25 and ‘once a day, every day’ and ‘more than once a day, every day’ = 0. In the composite 0–14 scale, 0 denotes consuming both sweets and sugared drinks at least once a day and 14 refers to never consuming sweets or sugary drinks. Indicators of ‘bottom-end’ health To identify adolescents who have substantially poorer health behaviours relative to their peers, we estimated the mean in the lower half of the distribution for each indicator separately, by country and year. Adolescents who fell below this threshold reported worse outcomes than an average respondent among the least well performing half of their population. This method has been applied in past HBSC,16 UNICEF17,18 and other international reports.27 The present study focuses on the individual determinants of scoring in the bottom-end group (separately by indicator, country and year) rather than on the country-level dispersion of health scores in the lower half of the distribution [see 27]. Analysis Our analysis focused on the strength of associations between the country-specific ridit-transformed FAS score and bottom-end health outcomes. We used linear probability models (i.e. ordinary least squares regressions with binary dependent variables) separately for each indicator, country and survey year. Such models are computationally straightforward and easy to interpret in terms of differences in probabilities of the outcome, but since they can produce biased estimates28 we also replicated all analyses using logistic regression. The findings were qualitatively identical (see the online appendix). All models controlled for age and gender and were fitted separately by country, using sample weights. Standard errors were adjusted to reflect the complex sampling structure of the surveys, with schools identified as primary sampling units and regions as strata (where relevant, i.e. in Belgium and the United Kingdom). We report on the effects of a one-unit difference in the ridit score for family affluence (the difference between the least and most affluent adolescents) on the probability of being in the bottom-end group for each indicator, controlling for age and gender. To investigate whether the effects of family affluence have changed significantly (P < 0.05) over the last four survey cycles, we included a trend variable (with 2001/2002 cycle as referent) with interaction terms between this trend variable and each of the three other predictors (i.e. family affluence, age and gender). Results There was substantial cross-national variation in the size of the bottom-end group of adolescents (i.e. those scoring below the mean of the lower half of the distribution in their country) to define the bottom-end group. The share of adolescents that ‘fell behind’ on MVPA ranged from 10% in Portugal to 27% in Belgium, while the proportion of those that ‘fell behind’ on healthy eating (fruit and vegetable consumption) varied from 16% in Belgium and Switzerland to 26% in Hungary and Spain. While one in four adolescents (25%) scored poorly on unhealthy eating (abstinence from sweets and sugary drinks) in the combined sample, this varied from 12% in Sweden to 33% in Norway. Table 1 shows large and widespread socio-economic inequalities in physical activity and healthy eating, with mixed results for unhealthy eating. In all 31 countries in the 2013/2014 analysis (note that there are 34 countries overall, but only 31 in 2013/2014), adolescents from more affluent households were significantly (P < 0.05) less likely to fall behind in MVPA, with the FAS gradient ranging from −8.3% in Finland (best performing) to −21.9% in Luxembourg (worst performing). Thus, the probability of scoring in the bottom-end group in MVPA is 21.9% points lower for adolescents from the most affluent families in Luxembourg than for their peers from the least affluent families. Similarly, adolescents in the lowest SES strata were more likely to fall further behind in healthy eating in all countries except Malta and Romania, where there were no statistically significant differences identified. The United Kingdom showed the largest negative gradient in healthy eating: the poorest adolescents were 21.3% points more likely to fall furthest behind. Hungary posted the largest negative gradient in unhealthy eating: adolescents from the most affluent families were 16.1% points less likely to consume excess sugar. In contrast, the most affluent adolescents in Estonia were 7.1% points more likely to engage in unhealthy eating. Table 1 Effect of family affluence on the probability of falling in the bottom-end group, controlling for age and gender (HBSC 2013/2014) Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Notes: Percentage point difference between the most and the least affluent adolescents; Standard errors adjusted for clustering at the school level and stratification by region (Belgium and the United Kingdom); Statistically significant effects (P < 0.05) in bold. Table 1 Effect of family affluence on the probability of falling in the bottom-end group, controlling for age and gender (HBSC 2013/2014) Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Country  Physical activity   Healthy eating   Unhealthy eating       B  95% CI  B  95% CI  B  95% CI  N  Austria  −0.111  (−0.158 to −0.063)  −0.106  (−0.158 to −0.054)  0.043  (−0.014 to 0.100)  3458  Belgium  −0.210  (−0.244 to −0.177)  −0.132  (−0.159 to −0.104)  −0.109  (−0.143 to −0.075)  10 285  Bulgaria  −0.151  (−0.203 to −0.100)  −0.144  (−0.188 to −0.100)  0.009  (−0.044 to 0.062)  4796  Canada  −0.134  (−0.174 to −0.094)  −0.203  (−0.248 to −0.157)  −0.027  (−0.068 to 0.014)  12 931  Croatia  −0.097  (−0.135 to −0.060)  −0.092  (−0.135 to −0.049)  0.030  (−0.015 to 0.074)  5741  Czech Republic  −0.123  (−0.160 to −0.085)  −0.130  (−0.173 to −0.087)  0.003  (−0.045 to 0.051)  5082  Denmark  −0.150  (−0.197 to −0.102)  −0.173  (−0.224 to −0.123)  −0.053  (−0.102 to −0.004)  3891  Estonia  −0.188  (−0.237 to to0.139)  −0.189  (−0.236 to −0.142)  0.071  (0.024 to 0.118)  4057  Finland  −0.083  (−0.113 to 0.053)  −0.145  (−0.184 to −0.106)  0.027  (−0.014 to 0.068)  5925  France  −0.136  (−0.177 to −0.095)  −0.143  (−0.187 to −0.099)  −0.072  (−0.124 to −0.019)  5636  Germany  −0.151  (−0.192 to −0.110)  −0.100  (−0.141 to −0.060)  −0.038  (−0.079 to 0.004)  5961  Greece  −0.086  (−0.129 to −0.044)  −0.124  (−0.167 to −0.081)  −0.021  (−0.063 to 0.022)  4141  Hungary  −0.119  (−0.156 to −0.083)  −0.176  (−0.226 to −0.125)  −0.161  (−0.221 to −0.101)  3935  Iceland  −0.135  (−0.166 to −0.104)  −0.128  (−0.159 to −0.098)  −0.037  (−0.067 to −0.007)  10 602  Ireland  −0.099  (−0.143 to −0.054)  −0.152  (−0.202 to −0.102)  −0.108  (−0.167 to −0.048)  4098  Italy  −0.158  (−0.207 to −0.109)  −0.155  (−0.203 to −0.107)  −0.034  (−0.091 to 0.023)  4072  Latvia  −0.216  (−0.254 to −0.178)  −0.150  (−0.190 to −0.110)  0.057  (0.016 to 0.099)  5557  Lithuania  −0.163  (−0.209 to −0.118)  −0.182  (−0.218 to −0.146)  −0.003  (−0.051 to 0.046)  5730  Luxembourg  −0.219  (−0.271 to −0.167)  −0.203  (−0.254 to −0.151)  0.000  (−0.055 to 0.056)  3318  Malta  −0.087  (−0.164 to −0.010)  −0.041  (−0.100 to 0.017)  0.024  (−0.021 to 0.070)  2265  The Netherlands  −0.201  (−0.242 to −0.161)  −0.164  (−0.210 to −0.117)  0.057  (0.009 to 0.104)  4301  Norway  −0.132  (−0.178 to −0.087)  −0.068  (−0.120 to −0.016)  0.039  (−0.016 to 0.094)  3072  Poland  −0.116  (−0.158 to −0.074)  −0.149  (−0.197 to −0.100)  −0.018  (−0.059 to 0.023)  4545  Portugal  −0.063  (−0.096 to −0.030)  −0.116  (−0.166 to −0.067)  −0.020  (−0.076 to 0.037)  4989  Romania  −0.129  (−0.173 to −0.085)  −0.030  (−0.082 to 0.022)  0.039  (−0.027 to 0.104)  3980  Slovakia  −0.170  (−0.208 to −0.132)  −0.128  (−0.164 to −0.092)  −0.042  (−0.085 to 0.001)  6099  Slovenia  −0.116  (−0.156 to −0.076)  −0.053  (−0.088 to −0.017)  0.012  (−0.035 to 0.058)  4997  Spain  −0.092  (−0.127 to −0.057)  −0.126  (−0.165 to −0.087)  −0.056  (−0.096 to −0.017)  11 136  Sweden  −0.203  (−0.238 to −0.168)  −0.104  (−0.135 to −0.073)  −0.024  (−0.051 to 0.004)  7700  Switzerland  −0.108  (−0.145 to −0.070)  −0.054  (−0.088 to −0.019)  0.036  (−0.002 to 0.074)  6634  United Kingdom  −0.192  (−0.218 to −0.166)  −0.213  (−0.240 to −0.186)  −0.067  (−0.096 to −0.037)  16 421  All  −0.145  (−0.152 to −0.137)  −0.139  (−0.147 to −0.131)  −0.022  (−0.030 to −0.014)  185 355  Notes: Percentage point difference between the most and the least affluent adolescents; Standard errors adjusted for clustering at the school level and stratification by region (Belgium and the United Kingdom); Statistically significant effects (P < 0.05) in bold. In the vast majority of the countries studied, the sizeable and statistically significant socio-economic gradient in MVPA remained stable over time (figure 1). In six countries, it became significantly (P < 0.05) larger in absolute terms between 2001/2002 and 2013/2014: Belgium, Italy, Latvia, the Netherlands, Sweden and the United Kingdom. Figure 1 View largeDownload slide FAS gradient in bottom-end physical activity between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Figure 1 View largeDownload slide FAS gradient in bottom-end physical activity between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 The FAS gradient in healthy eating remained stable over time in most of the countries studied (figure 2). In Canada and the United Kingdom this gradient increased, suggesting that adolescents from lower socio-economic backgrounds were increasingly less likely to consume fruit and vegetables when compared with their peers. In three other countries—Latvia, Lithuania and Romania—the FAS gradient has decreased. Although in Latvia and Lithuania adolescents from lower-affluence families were still significantly (P < 0.05) less likely to eat healthily, in Romania socio-economic inequalities were no longer statistically significant in 2013/2014. Figure 2 View largeDownload slide FAS gradient in bottom-end healthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Figure 2 View largeDownload slide FAS gradient in bottom-end healthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Conversely, in the majority of the countries analysed, there were no statistically significant socio-economic inequalities in unhealthy eating in 2013/2014. In eight countries, adolescents from more affluent households were less likely to report frequent consumption of sweets and sugary drinks, while in three others (Estonia, Latvia and the Netherlands), they were more likely to do so. While in many countries, there was not a significant (P < 0.05) FAS gradient in unhealthy eating between 2001/2002 and 2013/2014 (figure 3), higher consumption of sugar in less affluent households persisted over time in France, Ireland and the United Kingdom. Figure 3 View largeDownload slide FAS gradient in bottom-end unhealthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 Figure 3 View largeDownload slide FAS gradient in bottom-end unhealthy eating between 2002 and 2014 Source: HBSC 2001/2002, 2005/2006, 2009/2010, 2013/2014 The FAS gradient in unhealthy eating changed significantly (P < 0.05) between 2001/2002 and 2013/2014 in eight countries. In six of these—Estonia, Latvia, Lithuania, Poland, Portugal and Romania—adolescents from more affluent backgrounds were more likely to report a higher frequency of unhealthy eating than their peers in 2001/2002, but these differences either narrowed or disappeared by 2013/2014. In the remaining two countries, Belgium and Hungary, adolescents from less affluent backgrounds were increasingly more likely to report a higher frequency of unhealthy eating. Discussion Adolescence is a critical period of transition in the life course, characterized by biological, psychological and relational changes. Such transitions are of fundamental importance to health15 and set the stage for future patterns of adult inequalities.29 In the vast majority of studied countries between 2011/2012 and 2013/2014, adolescents from relatively low SES families had a greater likelihood of falling furthest behind in health, particularly with respect to reported physical activity and healthy eating. For unhealthy eating another pattern emerged, with no indication of an association with SES for the majority of the countries and mixed results for the rest. Sweets and soft drink consumption may be associated with knowledge about healthy diets, which is typically higher in more affluent socio-economic groups.30 Indeed, in eight countries (i.e. Belgium, Denmark, France, Hungary, Iceland, Ireland, Spain and the United Kingdom), young people from low socio-economic groups were overrepresented in the bottom-end of unhealthy eating in 2013/2014. In four of these (i.e. Belgium, France, Spain and the United Kingdom), the pattern has persisted since 2001/2002. Meanwhile, in another six countries (i.e. Croatia, Estonia, Latvia, Lithuania, Poland and Portugal) adolescents from more affluent backgrounds were more likely to report a higher frequency of unhealthy eating than their peers in 2001/2002, although the association decreased or disappeared by 2013/2014. This is consistent with historical studies that documented a positive association between family affluence and frequent soft drink consumption among adolescents in Central and Eastern Europe (CEE) at the turn of the century, possibly as an indicator of the ability to afford such luxuries and hence relative wealth.32 Inequalities in adolescent health were widening in some of the countries studied. Socio-economic inequalities in MVPA increased in six countries during the 12-year study period (Belgium, Italy, Latvia, the Netherlands, Sweden and the United Kingdom). In Canada and the United Kingdom, the association between SES and adolescent healthy eating became more pronounced over time, as did unhealthy eating in Belgium and Hungary. Widening national income inequalities tend to increase socio-economic differences in health, perpetuating the socio-economic divide.8 Conversely, in some (mostly CEE) countries, inequalities in adolescent health decreased considerably between 2001/2002 and 2013/2014. Adolescents in Latvia and Lithuania from less affluent families were less likely to eat healthily, but this association weakened significantly (P < 0.05) between 2001/2002 and 2013/2014. For Romania, socio-economic inequalities in healthy eating disappeared in the same period. These results suggest that social gradients in health can evolve over time. This study has several potential limitations. First, the cross-sectional design of the HBSC study precludes the establishment of the temporal sequence of events, limiting claims on causality. It is unlikely, however, that the outcomes of physical activity and diet would precede (or be determinants) of SES. Second, the indicators of adolescent health used here, although standardized and validated for cross-national comparison,26,33 come with their own inherent limitations. The indicators of fruit, vegetables, sweets and sugary drink consumption are all based on frequency of intake, rather than amounts consumed.2 Thus, we do not know whether, e.g. adolescents who reported daily consumption of fruit and vegetables actually consumed five portions of fruit and vegetables a day, but our measures are the best proxies for healthy and unhealthy eating afforded by our data. Our measure of SES relies on an assumption that ‘the scores of the [FAS] scale can be used to rank individuals and groups along a latent continuum of material wealth’.31 Although FAS has been validated against other measures of family SES, such as parental occupation,23 the scale is subject to measurement error; and it is very challenging to assess this construct by self-report in populations as young as 11 years. Finally, not all countries were present in each of the four survey cycles, limiting the cross-country comparability of the results. In conclusion, this study establishes that socio-economic inequalities in adolescent physical activity and dietary behaviours are large and stable across countries and over time. These are important findings because they suggest two common mechanisms—physical activity and diet—by which health inequalities emerge and adolescents become socially disadvantaged within and across different cultures and societies. The results indicate a fundamental unfairness16–18 that affects the most socio-economically disadvantaged adolescents in almost all societies; an unfairness that is consistent and persistent and sets the stage for negative health trajectories. These findings point to a universal need for public health interventions focused on physical activity and diet, including social policies that specifically target these and other aspects of health in our most disadvantaged children, as priorities, nationally and internationally. Acknowledgements The Health Behaviour in School-aged Children (HBSC) study is a World Health Organization collaborative study and is supported by each member country of the HBSC network (www.hbsc.org). The HBSC study is coordinated internationally by Dr Joanna Inchley, University of St. Andrews, Scotland, with international data coordination performed by Dr Oddrun Samdal, University of Bergen, Norway. Funding No external funding. Conflicts of interest: None declared. Key points Few cross-national studies have investigated the experiences of adolescents who report substantially worse health outcomes relative to others. This study reveals large, widespread and persistent socio-economic inequalities in the risks of the poorest adolescent health behaviours with respect to reported physical activity and healthy eating both across 34 industrialized countries and over time (between 2002 and 2014). Inequalities in unhealthy eating are less prevalent and may change their direction over time. References 1 Cavallo F, Dalmasso P, Ottová-Jordan V, et al.   Trends in self-rated health in European and North-American adolescents from 2002 to 2010 in 32 countries. Eur J Public Health  2015; 25: 13– 5. 2 Vereecken C, Pedersen TP, Ojala K, et al.   Fruit and vegetable consumption trends among adolescents from 2002 to 2010 in 33 countries. Eur J Public Health  2015; 25: 16– 9. http://dx.doi.org/10.1093/eurpub/ckv012 Google Scholar CrossRef Search ADS PubMed  3 Kalman M, Inchley J, Sigmundova D, et al.   Secular trends in moderate-to-vigorous physical activity in 32 countries from 2002 to 2010: a cross-national perspective. Eur J Public Health  2015; 25(suppl 2): 37– 40. Google Scholar CrossRef Search ADS PubMed  4 Currie C, Zanotti C, Morgan A, et al.   Social Determinants of Health and Well-Being Among Young People . Copenhagen: World Health Organization Regional Office for Europe, 2012. 5 de Looze M, Raaijmakers Q, Ter Bogt T, et al.   Decreases in adolescent weekly alcohol use in Europe and North America: evidence from 28 countries from 2002 to 2010. Eur J Public Health  2015; 25: 69– 72. http://dx.doi.org/10.1093/eurpub/ckv031 6 Hublet A, Bendtsen P, de Looze ME, et al.   Trends in the co-occurrence of tobacco and cannabis use in 15-year-olds from 2002 to 2010 in 28 countries of Europe and North America. Eur J Public Health  2015; 25: 73– 5. http://dx.doi.org/10.1093/eurpub/ckv032 7 Kuntsche E, Ravens-Sieberer U. Monitoring adolescent health behaviours and social determinants cross-nationally over more than a decade: introducing the Health Behaviour in School-aged Children (HBSC) study supplement on trends. Eur J Public Health  2015; 25: 1– 3. http://dx.doi.org/10.1093/eurpub/ckv009 Google Scholar CrossRef Search ADS PubMed  8 Elgar FJ, Pförtner T-K, Moor I, et al.   Socioeconomic inequalities in adolescent health 2002–2010: a time-series analysis of 34 countries participating in the Health Behaviour in School-aged Children study. The Lancet  2015; 385: 2088– 95. Google Scholar CrossRef Search ADS   9 Moor I, Rathmann K, Lenzi M, et al.   Socioeconomic inequalities in adolescent smoking across 35 countries: a multilevel analysis of the role of family, school and peers. Eur J Public Health  2015; cku244. 10 Magklara K, Skapinakis P, Niakas D, et al.   Socioeconomic inequalities in general and psychological health among adolescents: a cross-sectional study in senior high schools in Greece. Int J Equity Health  2010; 9: 3. http://dx.doi.org/10.1186/1475-9276-9-3 Google Scholar CrossRef Search ADS PubMed  11 Reiss F. Socioeconomic inequalities and mental health problems in children and adolescents: a systematic review. Soc Sci Med  2013; 90: 24– 31. http://dx.doi.org/10.1016/j.socscimed.2013.04.026 Google Scholar CrossRef Search ADS PubMed  12 Moor I, Lampert T, Rathmann K, et al.   Explaining educational inequalities in adolescent life satisfaction: do health behaviour and gender matter? Int J Public Health  2014; 59: 309– 17. Google Scholar CrossRef Search ADS PubMed  13 Frederick CB, Snellman K, Putnam RD. Increasing socioeconomic disparities in adolescent obesity. Proc Natl Acad Sci USA  2014; 111: 1338– 42. http://dx.doi.org/10.1073/pnas.1321355110 Google Scholar CrossRef Search ADS PubMed  14 Stalsberg R, Pedersen AV. Effects of socioeconomic status on the physical activity in adolescents: a systematic review of the evidence. Scand J Med Sci Sports  2010; 20: 368– 83. http://dx.doi.org/10.1111/j.1600-0838.2009.01047.x Google Scholar CrossRef Search ADS PubMed  15 Viner RM, Ozer EM, Denny S, et al.   Adolescence and the social determinants of health. The Lancet  2012; 379: 1641– 52. http://dx.doi.org/10.1016/S0140-6736(12)60149-4 Google Scholar CrossRef Search ADS   16 Currie C, Menchini L, Richardson D, et al.   Comparing inequality in the well-being of children in economically advanced countries: a methodology. Innocenti Work Paper:IWP-2010-19, 2011. 17 UNICEF. The Children Left behind: A League Table of Inequality in Child Well-Being in the World’s Rich Countries . Florence: UNICEF Innocenti Research Centre, 2010. 18 UNICEF. Fairness for Children: A league table of inequality in child well-being in rich countries, Innocenti Report Card 13. Florence: UNICEF Office of Research, 2016. 19 Heckman JJ. Skill formation and the economics of investing in disadvantaged children. Science  2006; 312: 1900– 2. http://dx.doi.org/10.1126/science.1128898 Google Scholar CrossRef Search ADS PubMed  20 McDaid D, Park A-L, Currie C, Zanotti C. Investing in the wellbeing of young people: making the economic case. In: McDaid D, Cooper C, editors. Economics of Wellbeing . Wellbeing: A Complete Reference Guide (Vol. 5). Oxford: Wiley-Blackwell, 2014, 181– 214. 21 Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: public-health crisis, common sense cure. The Lancet  2002; 360: 473– 82. http://dx.doi.org/10.1016/S0140-6736(02)09678-2 Google Scholar CrossRef Search ADS   22 Nishtar S, Gluckman P, Armstrong T. Ending childhood obesity: a time for action. The Lancet  2016; 387: 825– 7. http://dx.doi.org/10.1016/S0140-6736(16)00140-9 Google Scholar CrossRef Search ADS   23 Currie C, Molcho M, Boyce W, et al.   Researching health inequalities in adolescents: the development of the Health Behaviour in School-Aged Children (HBSC) family affluence scale. Soc Sci Med  2008; 66: 1429– 36. http://dx.doi.org/10.1016/j.socscimed.2007.11.024 Google Scholar CrossRef Search ADS PubMed  24 Currie C, Inchley J, Molcho M, et al.   Health Behaviour in School-Aged Children (HBSC) Study Protocol: Background, Methodology and Mandatory Items for the 2013/14 Survey . St Andrews: CAHRU, 2014. 25 WHO. Information sheet: Global recommendations on physical activity for health 5-17 years old [Internet]. World Health Organisation, 2011. Available at: http://www.who.int/dietphysicalactivity/publications/recommendations5_17years/en/ (11 March 2016, date last accessed). 26 Vereecken C, Rossi S, Giacchi MV, Maes L. Comparison of a short food-frequency questionnaire and derived indices with a seven-day diet record in Belgian and Italian children. Int J Public Health  2008; 53: 297– 305. Google Scholar PubMed  27 Chzhen Y, Bruckauf Z, Ng K, et al.   Inequalities in adolescent health and life satisfaction: Evidence form the Health Behaviour in School-Aged Children study. Florence: UNICEF Office of Research, 2016. Report no: 2016–11. 28 Horrace WC, Oaxaca RL. Results on the bias and inconsistency of ordinary least squares for the linear probability model. Econ Lett  2006; 90: 321– 7. http://dx.doi.org/10.1016/j.econlet.2005.08.024 Google Scholar CrossRef Search ADS   29 Patton GC, Coffey C, Cappa C, et al.   Health of the world’s adolescents: a synthesis of internationally comparable data. The Lancet  2012; 379: 1665– 75. Google Scholar CrossRef Search ADS   30 Hulshof K, Brussaard JH, Kruizinga AG, et al.   Socio-economic status, dietary intake and 10 y trends: the Dutch National Food Consumption Survey. Eur J Clin Nutr  2003; 57: 128– 37. http://dx.doi.org/10.1038/sj.ejcn.1601503 Google Scholar CrossRef Search ADS PubMed  31 Torsheim T, Currie C, Boyce W, et al.   Material deprivation and self-rated health: a multilevel study of adolescents from 22 European and North American countries. Soc Sci Med  2004; 59: 1– 12. http://dx.doi.org/10.1016/j.socscimed.2003.09.032 Google Scholar CrossRef Search ADS PubMed  32 Vereecken C, Inchley J, Subramanian SV, et al.   The relative influence of individual and contextual socio-economic status on consumption of fruit and soft drinks among adolescents in Europe. Eur. J Public Health  2005; 15: 224– 32. 33 Prochaska JJ, Sallis JF, Long B. A physical activity screening measure for use with adolescents in primary care. Arch Pediatr Adolesc Med  2001; 155: 554– 9. http://dx.doi.org/10.1001/archpedi.155.5.554 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.

Journal

The European Journal of Public HealthOxford University Press

Published: Jan 5, 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