. 2016 Jun 3;11(6):e0156945. doi: 10.1371/journal.pone.0156945

Table 3. Association between health outcomes and potential mediator.

Dependent variable	Independent variable	Coefficient / odds ratio (95% CI)	p
Step 2^a
Breakfast frequency	Maternal tertiary education	0.14 (0.05 to 0.23)	0.002
Breakfast frequency	Neighbourhood income	0.05 (-0.05 to 0.16)	0.31
Exercise frequency	Maternal tertiary education	0.14 (0.10 to 0.18)	< 0.001
Exercise frequency	Neighbourhood income	0.03 (-0.03 to 0.08)	0.32
Step 3^b
Obesity	Breakfast frequency	0.88 (0.84 to 0.92)	< 0.001
Obesity	Exercise frequency	0.86 (0.76 to 0.97)	0.01
Hypertension	Breakfast frequency	1.00 (0.97 to 1.03)	0.93
Hypertension	Exercise frequency	0.85 (0.78 to 0.92)	< 0.001

^a Regressing SES variables on mediators

^b Regressing mediators on health outcomes while controlling for maternal education level

All associations were estimated after the adjustment of child’s gender and age, paternal and maternal BMI, and intra-cluster correlation.