Table 2.

Change in BARI 2D Risk Factor Control Status

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.