Table 2. Estimates of Difference Between Treatment Arms in Developing Type 2 Diabetes.
| Analysis | INT-DPM vs INT | P value | INT vs CON | P value | INT-DPM vs CON | P value | Combined intervention vs CON | P value |
|---|---|---|---|---|---|---|---|---|
| OR unadjusted (95% CI) | 1.11 (0.7-1.65) | .59 | 0.53 (0.34-0.84) | .01 | 0.60 (0.3-0.93) | .02 | 0.57 (0.38-0.85) | .01 |
| OR adjusted (95% CI)a | 1.12 (0.75-1.65) | .59 | 0.54 (0.34-0.85) | .01 | 0.60 (0.38-0.94) | .02 | 0.57 (0.38-0.85) | .01 |
| OR adjusted (95% CI)b | 1.14 (0.77-1.70) | .51 | 0.54 (0.34-0.85) | .01 | 0.61 (0.39-0.96) | .03 | 0.57 (0.38-0.87) | .01 |
| HR unadjusted (95% CI) | 1.09 (0.76-1.57) | .63 | 0.53 (0.35-0.80) | .003 | 0.62 (0.41-0.92) | .02 | 0.57 (0.40-0.82) | .002 |
| HR adjusted (95% CI)c | 1.13 (0.78-1.63) | .51 | 0.53 (0.35-0.81) | .003 | 0.64 (0.43-0.97) | .03 | 0.58 (0.41-0.84) | .004 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CON, control arm without trial intervention. HR, hazard ratio; INT, standard intervention group; INT-DPM, intervention group with diabetes prevention mentors; OR, odds ratio.
a
Adjusted for duration of follow-up.
b
Adjusted for follow-up length and age, BMI, and fasting plasma glucose levels at baseline.
c
Adjusted at baseline for age, BMI, and fasting plasma glucose levels. The primary analysis in the Norfolk Diabetes Prevention Study was a logistic regression model with data presented as ORs. A secondary analysis using a proportional hazard model is also shown with HRs.