The pillars of health: influence of multiple lifestyle behaviors on body mass index and depressive symptoms in adult twins

Table 4 Unstandardized parameter estimates of depressive symptom score from the number of health pillars among same sex twins

Independent variables	All five health pillars		Two pillars (Sedentary time & Smoking)
	Men	Women	Men	Women
Phenotypic model
b_P	*-.120 (.013***)	*-.168 (.010***)	*-.146 (.014***)	*-.166 (.011***)
Quasi-causal model^a
b_P	*-.075 (.019***)	*-.112 (.015***)	*-.079 (.025***)	*-.127 (.020***)
b_A	*-.109 (.051***)	*-.133 (.037***)	*-.116 (.047***)	-.070 (.038)
b_C	*-.109 (.051***)	*-.133 (.037***)	*-.116 (.047***)	-.070 (.038)
Quasi-causal model^a,b
b_P	*-.098 (.012***)	*-.098 (.012***)	*-.110 (.015***)	*-.110 (.025***)
b_A	-.062 (.039)	*-.161 (.033***)	-.067 (.035)	*-.097 (.032***)
b_C	-.062 (.039)	*-.161 (.033***)	-.067 (.035)	*-.097 (.032***)

Bolded parameter estimates are statistically significant at p < .05. Depressive symptom score is square root transformed
b_P = phenotypic association between predictor and outcome. b_A = amount of variance in body mass index attributable to additive genetic influences. b_C = amount of variance in body mass index attributable to shared environmental influences
^a b_A and b_C are constrained to equality
^b b_P is constrained to be equal for men and women

ISSN: 1471-2458