Skip to main content

Table 4 Logistic regression analysis presenting adjusted odds ratios for each childhood health and well-being outcome by ACE count, resilience assets and demographics

From: Adverse childhood experiences and sources of childhood resilience: a retrospective study of their combined relationships with child health and educational attendance

  Common childhood conditions   Poor childhood healtha High school absenteeismb
Asthma Allergies Headaches Digestive
AOR (95%CIs) P AOR(95%CIs) P AOR(95%CIs) P AOR(95%CIs) P AOR(95%CIs) P AOR(95%CIs) P
ACE Count 0 Ref < 0.001   < 0.001   < 0.001   < 0.001   < 0.001   < 0.001
1 1.25 (0.89–1.75) 0.205 1.14 (0.86–1.52) 0.360 1.48 (1.14–1.92) 0.003 1.75 (1.26–2.44) < 0.001 1.41 (1.03–1.92) 0.031 2.04 (1.40–2.99) < 0.001
  2–3 1.91 (1.38–2.64) < 0.001 1.26 (0.94–1.69) 0.130 1.75 (1.33–2.28) < 0.001 1.56 (1.09–2.23) 0.015 1.79 (1.31–2.44) < 0.001 2.32 (1.59–3.39) < 0.001
  ≥4 2.76 (2.00–3.82) < 0.001 2.05 (1.51–2.77) < 0.001 2.71 (2.05–3.59) < 0.001 3.32 (2.36–4.67) < 0.001 2.59 (1.86–3.61) < 0.001 4.74 (3.26–6.89) < 0.001
Childhood resilience assets Community help   0.498   0.660   0.633   0.503   0.719   0.411
Adult available   0.760   0.838   0.572 0.73 (0.55–0.97) 0.030   0.182   0.595
Given Opportunities   0.051   0.583   0.992   0.571 0.64 (0.49–0.85) 0.002 0.57 (0.42–0.79) < 0.001
  Treated fairly   0.579 0.61 (0.46–0.81) < 0.001   0.188   0.436   0.276 0.60 (0.42–0.85) 0.004
  Culturally engaged   0.302   0.892   0.370   0.866   0.247   0.428
  Supportive friends   0.434   0.916 0.67 (0.51–0.88) 0.004   0.263 0.49 (0.36–0.67) < 0.001   0.123
  Role model   0.315   0.493   0.411   0.478 0.58 (0.42–0.79) < 0.001   0.114
ACE count* Childhood resilience assets Community help   0.241   0.886   0.306   0.322   0.274   0.530
Adult available   0.231   0.215   0.530   0.226   0.396   0.147
Given Opportunities   0.369   0.847   0.600   0.240   0.904   0.339
Treated fairly   0.526   0.926   0.155   0.291   0.480   0.479
Culturally engaged   0.686   0.811   0.317   0.833   0.967   0.213
Supportive friends   0.755   0.834   0.362   0.774   0.077   0.144
Role model   0.717   0.835   0.754   0.195   0.574   0.579
Age (years) 18–29 Ref < 0.001   0.046   < 0.001   0.641   0.002   0.042
30–39 0.83 (0.59–1.17) 0.284 0.99 (0.72–1.36) 0.931 0.71 (0.53–0.96) 0.025   0.265 0.73 (0.52–1.02) 0.068 1.08 (0.72–1.63) 0.697
  40–49 0.40 (0.28–0.59) < 0.001 0.73 (0.52–1.01) 0.055 0.68 (0.50–0.91) 0.010   0.904 0.57 (0.40–0.80) < 0.001 0.99 (0.66–1.48) 0.942
  50–59 0.48 (0.34–0.69) < 0.001 0.82 (0.60–1.14) 0.239 0.55 (0.40–0.74) < 0.001   0.215 0.51 (0.36–0.73) 0.000 1.15 (0.77–1.72) 0.484
  60–69 0.34 (0.23–0.51) < 0.001 0.66 (0.47–0.92) 0.013 0.52 (0.39–0.71) < 0.001   0.705 0.65 (0.46–0.91) 0.012 0.59 (0.37–0.93) 0.023
Sexc Male   0.233 0.79 (0.64–0.98) 0.030 0.68 (0.56–0.83) < 0.001 0.76 (0.59–0.98) 0.031   0.341   0.303
Deprivation quintile 1 (least deprived) Ref 0.991   0.232   0.677   0.236   0.144   < 0.001
2   0.701   0.163   0.677   0.395   0.189 1.44 (0.87–2.37) 0.153
  3   0.696   0.462   0.452   0.067   0.375 1.51 (0.93–2.43) 0.094
  4   0.834   0.981   0.590   0.210   0.106 2.55 (1.60–4.06) < 0.001
  5 (most deprived)   0.895   0.038   0.479   0.316   0.664 2.76 (1.72–4.44) < 0.001
Sampled Boost   0.685   0.289   0.695   0.174   0.269   0.723
Ethnicitye Other 0.32 (0.13–0.80) 0.015   0.608   0.553   0.271   0.525   0.056
  1. Ref = reference category. Logistic regression used a conditional model with only terms significantly contributing to the model’s fit to observed data (P < 0.05) retained in the final model. P values are shown for all terms entered into the model. AORs (Adjusted Odds Ratios) are only calculated for significant terms which remained in the model. Successive stages in model development for each outcome are shown in Additional file 1: Table S2. aSelf-rated overall health in childhood was categorised ‘poor’ for responses fair or poor; other options were excellent, very good or good; bSecondary school absenteeism was rated high if > 20 days per year; cReference group = female; dBoost is additional randomised sample from high Welsh language areas (> 40% Welsh speaking), reference group = core sample; eReference group = White