Table 4.
Hierarchical Regression Analyses Predicting HbA1c Levels in Older Adults (≥65 y) With Type 2 Diabetes (N = 430), the 1998 and 2000 Health and Retirement Study (HRS) and the HRS 2003 Diabetes Studya
| Variable | Model 1b | Model 2c | Model 3d | Model 4e | ||||
|---|---|---|---|---|---|---|---|---|
| β | P Value | β | P Value | β | P Value | β | P Value | |
| Demographics | ||||||||
| Race/ethnicity | ||||||||
| Non-Hispanic black | 0.13 | .49 | .17 | .37 | .09 | .32 | .11 | .56 |
| Hispanic | .24 | .25 | .27 | .20 | .24 | .23 | .25 | .21 |
| Non-Hispanic white | 1 [Reference] | 1 [Reference] | 1 [Reference] | 1 [Reference] | ||||
| Clinical Conditions | ||||||||
| No. of chronic diseases | — | — | −.03 | .45 | −.07 | .08 | −.06 | .13 |
| Duration of diabetes | — | — | .01 | .32 | −.01 | .32 | −.01 | .32 |
| Treatment Modality | ||||||||
| Diet only | — | — | — | — | −1.01 | <.001 | −1.07 | <.001 |
| Oral medication | — | — | — | — | −.46 | .001 | −.46 | .001 |
| Insulin only or combination | — | — | — | — | 1 [Reference] | 1 [Reference] | ||
| Lifestylef | ||||||||
| Fair | — | — | — | — | — | — | −.90 | .13 |
| Good | — | — | — | — | — | — | −.80 | .18 |
| Poor | — | — | — | — | — | — | 1 [Reference] | |
| Fg | .80 | .45 | 1.38 | .25 | 7.18 | <.001 | 5.71 | <.001 |
| R 2, % | .38 | — | 1.34 | — | 9.59 | — | 10.17 | — |
| ΔR 2, % | — | — | .96 | — | 8.25 | — | 0.58 | — |
| Fh | — | — | 2.07 | .13 | 19.30 | <.001 | 1.36 | .26 |
Abbreviation: HbA1c, hemoglobin A1c.
P values calculated using t test, analysis of variance (ANOVA), and simple regression; significance set at α = .05.
Model 1: demographic variables only.
Model 2: variables are demographics and clinical conditions.
Model 3: variables are demographics, clinical conditions, and treatment modality.
Model 4: variables are demographics, clinical conditions, treatment modality, and lifestyle.
Poor = 0 positive health behaviors, fair = 1 positive health behavior, good ≥2 positive health behaviors.
F test for model.
F test for change in R 2.