High school dropout and long-term sickness and disability in young adulthood: a prospective propensity score stratified cohort study (the Young-HUNT study)

Table 2 Risk difference ^* and odds ratio with 95% confidence intervals for long-term sickness or disability between age 24 to 29 years for high school dropouts versus school completers in the whole population and within the families

	Model 0	Model 1	Model 2	Model 3	Model 4	Model 5
Dropout *versus* completion (ref.)
Whole population
Risk difference	20.8 (17.0 to 24.7)	18.7 (15.0 to 22.4)	17.4 (13.8 to 21.1)	16.6 (13.0 to 20.4)	15.8 (12.2 to 19.5)	15.3 (11.7 to 19.0)
Odds ratio	3.92 (3.28 to 4.68)	3.53 (2.95 to 4.24)	3.34 (2.8 to 4.0)	3.20 (2.65 to 3.86)	3.07 (2.54 to 3.71)	2.96 (2.44 to 3.60)
Within family ¹
Odds ratio	1.89 (0.96 to 3.74)	–	2.03 (1.01 to 4.08)	2.53 (1.15 to 5.54)	2.48 (1.13 to 5.49)	2.39 (1.04 to 5.47)

*Estimated risk difference in the 6-year risk for long-term sickness and disability with the covariates at their mean.
Risk difference (in %, with 95% CI) and odds ratio (with 95% CI) in the whole population (logistic regression models, N=6612) and within the families (sibling fixed-effect models, N=316).
Model 0: adjusted for sex, age, and follow-up time.
Model 1: model 0 +adjusted for maternal education level.
Model 2: model 1 + adjusted for somatic disease, symptom load, psychological distress, concentration problems, insomnia, and self-rated health.
Model 3: model 2 + adjusted for overweight, smoking, and physical activity.
Model 4: model 3 + adjusted for self-esteem, subjective well-being, loneliness, and family living situation.
Model 5: model 4 + adjusted for reading and writing difficulties, bullying, disease-related school absence, educational aspirations, academic problems, school dissatisfaction, and school-related conduct.
¹In the Within-family models the covariate maternal education level is omitted.

ISSN: 1471-2458