Table 3.
Cause- Specific Hazard Ratios for Microsize MI and Typical MI according to Baseline Brain Natriuretic Peptide (NT-proBNP) Tertiles, Excluding Individuals with Possible Heart Failure (HF)
Events | Microsize MI | Events | Typical MI | P for equalitya | |
---|---|---|---|---|---|
Model 1 | |||||
< 42.5(Ref) | 10 | 1 | 55 | 1 | 0.05 |
42.5–115.5 | 25 | 3.24 (1.68–6.30) | 66 | 1.27 (0.87–1.87) | |
≥115.5 | 33 | 5.70 (2.65–12.29) | 61 | 2.46 (1.53–3.94) | |
Model 2 | |||||
< 42.50(Ref) | 10 | 1 | 55 | 1 | 0.06 |
42.5–115.5 | 25 | 3.17 (1.46–6.85) | 66 | 1.16 (0.70–1.92) | |
≥115.5 | 33 | 5.41 (2.14–13.69) | 61 | 2.15 (1.07–4.32) |
Possible heart failure (HF) defined as:
1. Baseline BNP > 125 if a participant is < than 75 years of age, or BNP > 450 if a participant ≥ 75 years
a. N = 362 among first sub-cohort (random sample, plus participants with CHD events)
b. N = 297 among second sub-cohort (random sample, plus participants with MI events)
2. Participants who had an incident MI event with concurrent prevalent HF (n = 18) and participants with MI and incident HF at the same date (n = 32)
3. Participants who had a first HF admission before an incident MI (n = 2)
Participants without MI event, but with incident HF before 01/01/2011 were censored at the time of first HF admission
Model 1 = BNP alone
Model 2 = Model 1 and Demographics (age, race, sex, household income, education, geographic region of residence), additional Framingham risk factors (current smoking status, systolic blood pressure, diabetes, HDL and total cholesterol) and other CVD risk factors and covariates (body mass index, log-transformed hsCRP, log-transformed ACR and medication use
Bold p < .05
aCause specific model using Lunn and McNeil approach for competing risk analyses