Genetic Influences on Individual Differences in Exercise Behavior during Adolescence

Table 4

Univariate model fitting results for exercise behavior in the three-age groups.

Model	vs	-2LL	df	𝜒^2	Δdf	𝑃	AIC
13-14 yr
(1) ACE: sex differences (rg estimated)	—	5482.577	2812	—	—	—	—
(2) ACE: sex differences (rg fixed at 0.5)	1	5482.704	2813	0.127	1	.72	−1.87
(3) ACE: no sex differences (rg fixed at 0.5)	2	5502.510	2815	19.81	2	<.01	13.93
(4)(a) CE: boys, ACE: girls (rg fixed at 0.5)	2	5536.641	2814	53.94	1	<.01	50.06
(4)(b) ACE: boys, CE: girls (rg fixed at 0.5)	2	5497.697	2814	14.99	1	<.01	11.12
(5)(a)AE: boys, ACE: girls (rg fixed at 0.5)	2	𝟓𝟒𝟖𝟑.𝟐𝟐𝟑	𝟐𝟖𝟏𝟒	𝟎.𝟓𝟐	𝟏	.𝟒𝟕	−𝟑.𝟑𝟓
(5)(b)ACE: boys, AE: girls (rg fixed at 0.5)	2	5502.504	2814	19.80	1	<.01	15.93
15-16 yr
(1) ACE: sex differences (rg estimated)	—	5943.005	2986	—	—	—	—
(2) ACE: sex differences (rg fixed at 0.5)	1	5944.573	2987	1.57	1	.21	−.43
(3) ACE: no sex differences (rg fixed at 0.5)	2	5949.728	2989	5.16	2	.08	.72
(4) CE: no sex differences (rg fixed at 0.5)	3	6024.535	2990	74.81	1	<.01	73.53
(5) AE: no sex differences (rg fixed at 0.5)	3	𝟓𝟗𝟓𝟎.𝟔𝟕𝟒	𝟐𝟗𝟗𝟎	.𝟗𝟓	𝟏	.𝟑𝟑	−𝟐.𝟑𝟑
17–19 yr
(1) ACE: sex differences (rg estimated)	—	4455.979	2158	—	—	—	—
(2) ACE: sex differences (rg fixed at 0.5)	1	4455.979	2159	.00	1	>.99	−2.00
(3) ACE: no sex differences (rg fixed at 0.5)	2	4458.120	2161	2.14	2	.34	−3.86
(4) CE: no sex differences (rg fixed at 0.5)	3	4495.737	2162	37.62	1	<.01	21.76
(5) AE: no sex differences (rg fixed at 0.5)	3	𝟒𝟒𝟓𝟖.𝟏𝟐𝟎	𝟐𝟏𝟔𝟐	.𝟎𝟎	𝟏	> .99	−𝟓.𝟖𝟔

Note. vs: versus; -2LL: −2log likelihood; df = degrees of freedom; 𝜒² = chi-square test statistic; Δdf = degrees of freedom of 𝜒² test; 𝑃=𝑃-value; AIC = Akaike’s Information Criterion; rg = genetic correlation between DOS twins. Most parsimonious models are printed in boldface type.