Skip to main content
. 2017 Nov 12;7(11):e018264. doi: 10.1136/bmjopen-2017-018264

Table 2.

Cross-sectional adjusted associations between time of day (fitted as continuous variable) and cardiovascular disease (CVD) risk factors measured in the British Regional Heart Study (BRHS) men (aged 60–79) attending the follow-up year 20 examination in 1998–2000

CVD risk factor‡ Model 1: age adjusted* Model 2: fully adjusted† Model 3: fully adjusted† excluding men with diabetes
Per cent difference (95% CI) in the CVD risk factor levels per hour of sampling§ p Value Per cent difference (95% CI) in the CVD risk factor levels per hour of sampling§ p Value Per cent difference (95% CI) in the CVD risk factor levels per hour of sampling p Value
NT-proBNP 3.5 (2.0 to 5.0) <0.001 3.3 (1.9 to 4.8) <0.001 3.5 (2.0 to 5.0) <0.001
IL-6 2.6 (1.7 to 3.4) <0.001 2.6 (1.8 to 3.4) <0.001 2.4 (1.6 to 3.3) <0.001
t-PA −3.3 (-3.8 to −2.9) <0.001 −3.3 (−3.7 to −2.9) <0.001 −3.2 (−3.6 to −2.7) <0.001
Fibrinogen −0.2 (0.5 to 0.0) 0.088 −0.2 (0.5 to 0.1) 0.104 −0.2 (0.5 to 0.1) 0.149
cTnT −0.4 (0.9 to 0.2) 0.194 −0.4 (1.0 to 0.2) 0.174 −0.4 (1.0 to 0.2) 0.165
CRP −1.0 (2.3 to 0.4) 0.151 −0.9 (2.2 to 0.4) 0.175 −0.9 (2.2 to 0.5) 0.191
vWF −0.2 (0.6 to 0.2) 0.374 −0.2 (0.6 to 0.2) 0.380 −0.1 (0.5 to 0.3) 0.703
D-dimer −0.0 (1.0 to 1.0) 0.929 −0.1 (1.0 to 0.9) 0.890 −0.1 (1.2 to 0.9) 0.801

Associations are reported as per cent difference in CVD risk factors levels per 1 hour of sampling over the examination day (08:00–19:00 hours). The statistically significant associations are marked in bold

*Model 1: two level linear models (level 1=person and level 2=town of residence during the BRHS recruitment) adjusted for age. Model 1 used the same number of observations of Model 2 (complete case analysis).

†Model 1 additionally adjusted for social class, body mass index, smoking status, alcohol consumption, physical activity, use of statin and season. Associations with IL-6, t-PA, fibrinogen, CRP, vWF and D-dimer were additionally adjusted for prevalence of stroke/MI, while association of time of the day with NT-proBNP and cTnT models were additionally adjusted for prevalence of heart failure.

‡Model 1 and Model 2 used the same number of observations: 3580 for NT-proBNP, 3832 for IL-6, 3863 for t-PA, 3861 for fibrinogen, 3827 for cTnT, 3838 for CRP, 3863 for vWF, 3859 for D-dimer.

§Model 3 number of observations: 3176 for NT-proBNP, 3398 for IL-6, 3429 for t-PA, 3427 for fibrinogen, 3398 for cTnT, 3405 for CRP, 3429 for vWF, 3425 for D-dimer.

CRP, C reactive protein; cTnT, cardiac troponin T; IL-6, interleukin-6; MI, myocardial infarction; NT-proBNP, N-terminal pro-brain natriuretic peptide; t-PA, tissue plasminogen activator; vWF, von Willebrand factor.