Table 2.
Risk Factor | Percent of Participants In Control† | Jump* p-val |
Slope* p-val |
|||
---|---|---|---|---|---|---|
Baseline | Year 1 | Year 3 | Year 5 | |||
LDL-C | 60 | 75 | 83 | 85 | <0.0001 Jump |
<0.0001 Improve |
Non-HDL-C | 54 | 71 | 79 | 83 | <0.0001 Jump |
<0.0001 Improve |
Triglycerides | 51 | 57 | 61 | 64 | 0.005 Jump |
<0.0001 Improve |
Systolic BP | 50 | 56 | 62 | 62 | 0.0013 Jump |
<0.0001 Improve |
Diastolic BP | 69 | 71 | 75 | 79 | 0.64 No Jump |
<0.0001 Improve |
Non-Smoker | 88 | 90 | 91 | 91 | 0.0013 Jump |
0.12 Maintain |
A1C | 40 | 51 | 48 | 46 | <0.0001 Jump |
<0.0001 Degrade |
Meet all goals | 6 | 10 | 13 | 13 | <0.0001 Jump |
0.13 Maintain |
p-values computed using a GEE logit(meet goal) = B0+B1*baseline +B2* year, taking into account multiple observations per participant. Baseline is an indicator variable for the baseline value and year is continuous year of follow-up (0–5). Testing the significance of B1 is testing if the baseline value can be predicted by the same slope as the year 1–5 data. The jump p-value corresponds to the test B1=0. The slope p-value corresponds to the test B2=0. Significance tests based on N=1854 participants with complete data at each of baseline, year 1 and year 3.
N=1854. These include three year survivors who had complete data at each of the baseline, year 1 and year 3 visits. At year 5, N=1062 due to late recruitment, loss to follow-up and deaths. Analysis restricted to reflect changes in risk factor status of individuals across time. Allowing all data to be used (not restricting to 3-year survivors with baseline, year 1 and year 3 visits) the percentages in the above table change less than 2% and the conclusion from the tests remain the same, with all significant p-values < 0.001 and all non-significant p-values > 0.075.