Table 2.

Co-twin control analyses evaluating the association between Adult Antisocial Behavior (AAB) Extended Symptom Counts at age 29 and Marital Status at age 29.

Outcome	Fixed Effect Estimates (SE)
	Between-pair		Within-pair		Prior AAB
	MZ	DZ	MZ	DZ	Age 17	Age 20
AAB at 29
1) No covariate model	−.57 (.15)***	−.57 (.23)***	−.26 (.14)**	−.29 (.17)**	---	---
2) Controlling for AAB at age 20	−.36 (.14)***	−.55 (.21)***	−.24 (.15)*	−.25 (.18)*	---	.68 (.07)***
3) Controlling for AAB at age 17	−.43 (.14)***	−.46 (.21)***	−.18 (.14)*	−.31 (.16)*	.32 (.04)***	---

Note. MZ and DZ represented monozygotic and dizygotic twin pairs, respectively. The former share 100% of their genetic material, whereas the latter share, on average, 50% of their segregating genetic material. Results from the first or no covariate model are also presented in Figure 2. To facilitate interpretation of these unstandized fixed effect estimates, the AAB extended symptom count variable was standardized to have a mean of 0 and a standard deviation of 1 prior to analysis. Between-pair effects can be used to approximate the individual-level effects. Significant within-pair estimates for MZ twins are indicative of a non-shared environmentally-mediated relationship between marriage and AAB, and particularly so when the DZ within-pair estimate is equivalent to the MZ estimate. Because AAB decreases with the advent of marriage, both the between- and within-pair effect estimates are negative. The prior AAB covariates, by contrast, are positive, since higher levels of AAB at ages 17 or 20 predict higher levels of AAB at age 29.

^***indicates that the fixed effect estimate is statistically significant at p < .001.

** and * indicate that the fixed effect estimate is statistically significant at p < .01 and p < .05, respectively.