Table 3.

Summary of covariate effects on latent change in CES-D and HbA_1c in full sample

	Latent slope		9 months		15 months
	CES-D	HbA1c	CES-D	HbA1c	CES-D	HbA1c
Site	−0.04 (0.10)	−0.01 (0.02)	0.92 (1.17)	0.32 (0.16)*	−0.25 (0.20)	0.65 (1.43)
Private insurance, comparison: public insurance	−0.97 (0.93)	−0.03 (0.06)*	0.46 (1.46)	−0.58 (0.20)**	−0.80 (0.25)**	−0.69 (1.79)
Parent race, comparison: non-Hispanic White
Hispanic	−0.13 (0.12)	−0.03 (0.02)	0.90 (1.64)	−0.30 (0.32)	−0.45 (0.39)	0.13 (1.51)
Non-Hispanic Black	0.14 (0.15)	0.003 (0.03)	−0.80 (1.59)	1.23 (0.30)**	1.25 (0.31)***	0.05 (2.24)
All other race/ethnicity groups	−0.08 (0.15)	0.01 (0.02)	−2.82 (1.82)	0.23 (0.23)	0.26 (0.32)	−3.28 (2.22)
DKA at diagnosis	−0.05 (0.08)	−0.03 (0.01)*	1.05 (0.93)	0.17 (0.13)	−0.01 (0.14)	0.78 (1.14)

Models for 9-month and 15-month outcomes centered on the latent intercept at the specific assessment. Consequently, the effects reflect covariate effects on latent mean outcome for each end point. Latent slope effects were equivalent in each model and can be interpreted in latent change in each outcome per month. β (SE) reported in each cell. “All other race/ethnicity groups” category includes Asian/Asian American, multiracial, and American Indian/Alaska Native participants.

^*P < 0.05;

^**P < 0.01;

^***P < 0.001.