Table 3.
Multivariable-adjusted associations of clinical predictors with longitudinal changes in LV diastolic function traits
| Predictor | Change in E/A | Change in E′ (cm/s) | Change in E/E′ | |||
|---|---|---|---|---|---|---|
| Est beta ± SE | P value | Est beta ± SE | P value | Est beta ± SE | P value | |
| Age | −0.02±0.01 | 0.005 | −0.30±0.04 | <0.0001 | 0.35±0.05 | <0.0001 |
| Female sex | −0.01±0.01 | 0.36 | −0.51±0.07 | <0.0001 | 0.70±0.09 | <0.0001 |
| Systolic blood pressure | −0.01±0.01 | 0.18 | −0.11±0.05 | 0.02 | 0.11±0.06 | 0.06 |
| Δ Systolic blood pressure | -- | -- | −0.12±0.04 | 0.006 | 0.22±0.05 | <0.0001 |
| Diastolic blood pressure | −0.02±0.01 | 0.05 | −0.10±0.05 | 0.03 | 0.05±0.05 | 0.28 |
| Δ Diastolic blood pressure | −0.02±0.01 | 0.004 | −0.13±0.05 | 0.005 | -- | -- |
| BMI | 0.00±0.01 | 0.94 | −0.07±0.04 | 0.06 | 0.05±0.05 | 0.26 |
| Δ BMI | -- | -- | −0.13±0.03 | 0.0002 | -- | -- |
| Total/HDL cholesterol | 0.01±0.01 | 0.51 | −0.02±0.05 | 0.63 | −0.05±0.06 | 0.39 |
| Δ Total/HDL cholesterol | -- | -- | -- | -- | -- | -- |
| ln(Triglycerides) | −0.02±0.01 | 0.02 | −0.01±0.05 | 0.88 | 0.06±0.06 | 0.32 |
| Δ ln(Triglycerides) | −0.03±0.01 | <0.0001 | -- | -- | -- | -- |
| Heart rate | −0.04±0.01 | <0.0001 | −0.12±0.04 | 0.004 | −0.03±0.04 | 0.50 |
| Δ Heart rate | −0.05±0.01 | <0.0001 | −0.17±0.04 | <0.0001 | -- | -- |
| Hypertension treatment status | −0.00±0.01 | 0.97 | 0.02±0.08 | 0.75 | −0.11±0.09 | 0.22 |
| Smoking | −0.03±0.02 | 0.13 | 0.03±0.13 | 0.79 | −0.21±0.16 | 0.21 |
| Diabetes | 0.00±0.02 | 0.86 | −0.35±0.12 | 0.003 | 0.36±0.15 | 0.01 |
The estimated beta coefficients represent the estimated change in the echocardiographic trait for each 1-SD higher value of the predictor variable.
Δ indicates continuous ‘change variables’ which are defined as the follow-up value minus the baseline value
Each 1 SD is equal to: 7.9 years for age, 18.7 mm Hg for systolic blood pressure, 16.4 mm Hg for change in systolic blood pressure, 8.5 mm Hg for diastolic blood pressure, 8.4 mm Hg for change in diastolic blood pressure, 5.3 kg/m2 for BMI, 2.1 kg/m2 for change in BMI, 1.0 for total/HDL cholesterol, 0.8 for change in total/HDL cholesterol, 0.5 for ln(triglycerides), 0.4 for change in ln(triglycerides), 9.5 beats per minute for heart rate, 8.5 beats per minute for change in heart rate.
The multivariable models are additionally adjusted for the interval between the 2 examination cycles and the baseline value of the echocardiographic trait. Fasting glucose, change in fasting glucose, change in hypertension treatment status, change in smoking, and change in diabetes were included as potential predictor variables but did not meet criteria for model inclusion.