Factors affecting children’s mental health and wellbeing: findings

Footnotes

1 The Scottish Government, “National Performance Framework,” [Online]. Available: https://nationalperformance.gov.scot/.

2 The Scottish Government, “Getting it right for every child (GIRFEC),” [Online]. Available: https://www.gov.scot/policies/girfec/.

3 The Scottish Parliament, “Children and Young People (Scotland) Act 2014,” Acts of the Scottish Parliament, 2014. [Online]. Available: http://www.legislation.gov.uk/asp/2014/8/contents/enacted.

4 Parkinson, J., Establishing a core set of national, sustainable mental health indicators for children and young people in Scotland: Final Report. 2012, NHS Health Scotland.

5 Parkinson, J., Establishing a core set of national, sustainable mental health indicators for children and young people in Scotland: Final Report. 2012, NHS Health Scotland.

6 Scottish Government, Scottish Schools Adolescent Lifestyle and Substance Use Survey (SALSUS) 2015: Mental Wellbeing Report. 2017: Edinburgh.

7 Goodman, R., H. Meltzer, and V. Bailey, The strengths and difficulties questionnaire: A pilot study on the validity of the self-report version.

8 The ‘very high’ group was selected as the outcome measure of interest for emotional problems and conduct problems because rates of emotional problems were high in our sample (17% of all S1-S4 pupils and 34% of S4 girls scored as ‘very high’). Widening the outcome to include ‘high’ or above, or ‘slightly raised’ or above, would have resulted in a highly prevalent outcome and less meaningful conclusions.

9 See the Realigning Children’s Services Technical report for the Wellbeing Survey Programme, by ScotCen, for full details of measures and cut-offs.

10 Parkinson, J., Establishing a core set of national, sustainable mental health indicators for children and young people in Scotland: Final Report. 2012, NHS Health Scotland.

11 Huppert, F.A., et al., The science of well-being. 2005, Oxford: Oxford University Press.

12 Clarke, A., et al., Warwick-Edinburgh Mental Well-being Scale (WEMWBS): validated for teenage school students in England and Scotland. A mixed methods assessment. BMC public health, 2011. 11(1): p. 487-487.

13 See the Realigning Children’s Services Technical report for the Wellbeing Survey Programme, by ScotCen, for full details of measures and cut-offs.

14 Huppert, F.A., et al., The science of well-being. 2005, Oxford: Oxford University Press.

15 Huebner, E.S., Initial Development of the Student's Life Satisfaction Scale. School Psychology International, 1991. 12(3): p. 231-240.

16 See the Realigning Children’s Services Technical report for the Wellbeing Survey Programme, by ScotCen, for full details of measures and cut-offs.

17 The analysis only examined the total number of risk factors in each domain for the secondary school data (since the primary school survey had different numbers of questions for each domain).

18 R Development Core Team, R: A language and environment for statistical computing. 2010, R Foundation for Statistical Computing: Vienna, Austria.

19 Multiple imputation is a statistical tool that creates multiple datasets where missing values are imputed (assigned plausible values) based on the other available data for a participant. Analysis is carried out on each created dataset and averaged across, to account for the uncertainty in imputed values.

20 Exact odds ratios and significance levels for these analyses can be made available upon request.

21 See appendix for the exact survey questions and response options used to identify each risk factor (including details of which types of prejudice were included etc.)

22 The analysis did not examine the total number of risk factors in the Health domain, since this only included two indicators.

23 Figure 5 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each individual risk factor, predicting the odds of having very high emotional problems, controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

24 Figure 6 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each domain, predicting the odds of having very high emotional problems based on the number of risk factors in that domain and controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

25 Table 2 summarises the results of one binomial logistic regression model for very high emotional problems with multiple predictors (the number of risk factors in each domain) controlling for child characteristics listed in Section 2.4.1. Predictors labelled as ‘still significant’ had p<.05 in the multivariable model.

26 Figure 7 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each individual risk factor, predicting the odds of having very high conduct problems, controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

27 Figure 8 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each domain, predicting the odds of having very high conduct problems based on the number of risk factors in that domain and controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

28 Table 4 summarises the results of one binomial logistic regression model for very high conduct problems, with multiple predictors (the number of risk factors in each domain) controlling for child characteristics listed in Section 2.4.1. Predictors labelled as ‘still significant’ had p<.05 in the multivariable model.

29 Figure 9 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each individual protective factor, predicting the odds of having high positive mental wellbeing controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

30 Figure 10 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each domain, predicting the odds of having high positive mental wellbeing based on the number of protective factors in that domain and controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

31 Table 5 summarises the results of one binomial logistic regression model for high positive mental wellbeing, with multiple predictors (the number of protective factors in each domain) controlling for child characteristics listed in Section 2.4.1. Predictors labelled as ‘still significant’ had p<.05 in the multivariable model.

32 Exact odds ratios and significance levels for these analyses can be made available upon request.

33 See the Realigning Children’s Services Technical report for the Wellbeing Survey Programme, by ScotCen, for full details of measures and cut-offs.

34 Figure 13 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each individual risk factor, predicting the odds of having low mood, controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

35 ‘Examining factors together’ refers to one binomial logistic regression model for low mood with all risk factors as predictors, controlling for child characteristics listed in Section 2.4.1.

36 Figure 14 illustrates adjusted relative risks (based on odds ratios) from separate binomial logistic regression models for each individual protective factor, predicting the odds of having high life satisfaction, controlling for child characteristics listed in Section 2.4.1. Adjusted relative risk values have been rounded for illustrative purposes.

37 Examining associations ‘simultaneously’ refers to one binomial logistic regression model for high life satisfaction with all risk factors as predictors, controlling for child characteristics listed in Section 2.4.1.

38 Scottish Government, Scottish Schools Adolescent Lifestyle and Substance Use Survey (SALSUS) 2015: Mental Wellbeing Report. 2017: Edinburgh.

39 Scottish Government, Scottish Schools Adolescent Lifestyle and Substance Use Survey (SALSUS) 2015: Mental Wellbeing Report. 2017: Edinburgh.

40 Salomon, I. and C.S. Brown, The Selfie Generation: Examining the Relationship Between Social Media Use and Early Adolescent Body Image. The Journal of Early Adolescence. 0(0): p. 0272431618770809.