Table 2:

Full model: Selected Predictors from the Lasso Regression for Mania/hypomania in Young Adulthood

Variable	Lasso exp(B)	p-valuea
Age at BD Onset	.	.
Age at Mania/hypomania Onset	.	.
Oldest Age (<18) of Mania/hypomania	1.62	<0.001
BD Subtype (BDI vs. BDII)	.	.
>1 manic/hypomanic episode (Y/n)	.	.
Depression Under 18 (Y/n)	.	.
Demographics: Sex (M/F)	.	.
Demographics: SES	.	.
Demographics: Race (white/non-white)	.	.
Family History: Depression (Y/n)	.	.
Family History: Mania (Y/n)	1.99	<0.001
Family History: Suicide Attempt (Y/n)	1.15	0.01
Family History: Psychosis (Y/n)	.	.
Family History: Anxiety (Y/n)	.	.
Family History: SUD (Y/n)	.	.
Family History: ADHD (Y/n)	.	.
Family History: Conduct Disorder (Y/n)	.	.
Comorbidity: Anxiety Under 18 (Y/n)	.	.
Comorbidity: ADHD Under 18 (Y/n)	.	.
Comorbidity: DBD Under 18 (Y/n)	.	.
Comorbidity: SUD Under 18 (Y/n)	.	.
Nuisance: Intake Age	1.08	0.07
Nuisance: Length of Adult Follow-up	.	.
Nuisance: # Adult Follow-up Visits	1.06	0.05

^ap-values calculated by fitting an unpenalized logistic regression model using lasso-selected predictors