Table 2.
Results of linear mixed models of direct and diabetes distress–buffering effects of autonomy support on subsequent 12-month HbA1c percent
| Direct-effect model |
Distress-buffering model |
|||||
|---|---|---|---|---|---|---|
| B | SE | P value | B | SE | P value | |
| Predictor | ||||||
| Autonomy support | −0.16 | 0.07 | 0.023 | −0.14 | 0.08 | 0.064 |
| Diabetes distress | 0.12 | 0.05 | 0.032 | 0.11 | 0.05 | 0.029 |
| Autonomy support × diabetes distress | −0.13 | 0.06 | 0.027 | |||
| Control variable | ||||||
| Age | −0.01 | 0.01 | 0.856 | −0.01 | 0.01 | 0.878 |
| Race/ethnicity* | 0.17 | 0.15 | 0.248 | 0.18 | 0.15 | 0.229 |
| Insulin use† | 0.39 | 0.12 | 0.001 | 0.38 | 0.12 | 0.002 |
| Time‡ | 0.01 | 0.01 | 0.972 | 0.01 | 0.01 | 0.961 |
| Prior average HbA1c (%) | 0.53 | 0.04 | <0.001 | 0.52 | 0.04 | <0.001 |
All coefficients are unstandardized and represent fixed effects. Boldface indicates significance at P < 0.05. The sample included 734 HbA1c values clustered within 308 individuals.
*Non-Hispanic white vs. other race and/or ethnicity.
†No insulin use = 0 vs. insulin use = 1.
‡Months between assessment of predictor variables (i.e., diabetes distress, autonomy support) and each HbA1c measurement.