The baseline characteristics for the total sample and the lowest and highest quartile groups of PF volume are presented in Table 1.
Distribution and determinants of the PF volume
Median (IQR) PF volume in the total cohort were 1.411 (IQR 1.035, 2.057) dl. Since males have a higher PF volume than females (median 1.729 dl, IQR 1.202,2.492; median 1.215 dl, IQR 0.909,1.552; respectively), the upper and lower PF volume quartiles of males and females were combined for the analysis (Figure 3A).
Fig. 3. The variation of PF volume to sex, age and BMI in a healthy population. PF volume is higher in males as in females (A), PF volume is not related to age (B) and PF volume is associated with BMI (C).
There was a significant difference between the lowest and highest quartile groups for age (55.7 ± 8.0 versus 59.1 ± 7.4, p = 0.015), BMI (23.7 ± 2.7 versus 28.1 ± 2.9, p value <0.001), glucose (5.5 ± 0.8 versus 5.9 ± 1.2, p = 0.025), HDL cholesterol (1.5 ± 0.4 versus 1.2 ± 0.4) and triglycerides (1.5 ± 1.1 versus 2.4 ± 2.0, p = 0.001), see table 1. The CAD findings were not different between the two groups of high and low PF. However, Framingham Risk Score was higher in the high PF group, possibly due to the association of PF with age and BMI.
Table 1. Baseline characteristics of the study sample, and divided into highest and lowest quartiles of PF.
|
Total sample (n=254)
|
PF low (n=65)
|
PF high (n=65)
|
P-value
|
Demographics
|
Age (years)
|
57.0 ± 7.5
|
55.7 ± 8.0
|
59.1 ± 7.4
|
0.015
|
Sex (% female)
|
48
|
46
|
48
|
0.860
|
Cardiovascular risk factors
|
Framingham Risk Score
Glucose (mmol/L)
|
18.0 ± 13.2
5.6 ± 0.9
|
14.4 ± 10.1
5.5 ± 0.8
|
21.4 ± 16.1
5.9 ± 1.2
|
0.004
0.025
|
Body mass index (kg/m2)
|
26.4 ± 3.7
|
23.7 ± 2.7
|
28.1 ± 2.9
|
<0.001
|
Systolic bloodpressure (mmHg)
|
142 ± 20
|
141 ± 23
|
146 ± 20
|
0.139
|
Diastolic bloodpressure (mmHg)
|
81 ± 11
|
80 ± 12
|
82 ± 11
|
0.254
|
Total cholesterol (mmol/L)
|
5.6 ± 1.1
|
5.5 ± 1.2
|
5.8 ± 1.2
|
0.148
|
HDL cholesterol (mmol/L)
|
1.3 ± 0.4
|
1.5 ± 0.4
|
1.2 ± 0.4
|
0.001
|
LDL cholesterol (mmol/L)
|
3.6 ± 1.0
|
3.4 ± 1.0
|
3.6 ± 1.1
|
0.405
|
Triglycerides (mmol/L)
|
1.5 (1.0, 2.2)
|
1.2 (0.8, 1.5)
|
1.7 (1.3, 2.5)
|
<0.001
|
Creatinine (μmol/L)
|
76 ± 17
|
76 ± 15
|
75 ± 18
|
0.769
|
eGFR (MDRD) (ml/min/1.73m2)
|
88 ± 18
|
89 ± 16
|
90 ± 21
|
0.619
|
CRP (mg/L)
Coronary Artery Disease
No Plaque (%)
Mild (%)
Moderate (%)
Severe (%)
Multi-vessel (%)
|
2.3 ± 2.7
39.4 ± 4.9
37.0 ± 4.8
10.20 ± 3.0
11.8 ± 3.2
1.6 ± 1.3
|
2.1 ± 2.5
46.2 ± 5.0
33.8 ± 4.8
7.7 ± 2.7
9.2 ± 2.9
3.1 ± 1.7
|
2.8 ± 3.8
35.4 ± 4.8
36.9 ± 4.9
10.8 ± 3.1
15.4 ± 3.6
1.5 ± 1.2
|
0.470
0.215
0.716
0.548
0.289
0.563
|
Data are presented as means ± standard deviation, percentage, or as median (interquartile range, IQR).
Distribution and determinants of diastolic function
The association between diastolic function and PF volume was investigated, as some of the diastolic parameters are expected to deteriorate during the development of diastolic dysfunction before clinical criteria for diastolic dysfunction are met (Figure 4).
Fig. 4. PF is not associated with diastolic function parameters in a healthy population. Data of the entire cohort (n=254) are displayed. No correlations are found.
Table 2. Cardiac function measured by transthoracic echocardiography.
|
Total population (n=254)
|
PF low (n=65)
|
PF high (n=65)
|
P-value
|
|
Left ventricular ejection fraction (%)
Left ventricular mass index (g/m2)
|
61 ± 5
84.7 ± 16.9
|
62 ± 5
80.6 ± 15.6
|
61 ± 5
88.0 ± 16.0
|
0.213
0.008
|
Left atrial volume index (ml/m)
|
33.7 ± 0.7
|
36.8 ± 10.3
|
32.7 ± 8.4
|
0.015
|
e’ lateral (cm/s)
|
11.0 ± 2.7
|
12.2 ± 2.9
|
10.3 ± 2.0
|
0.005
|
e’ septal (cm/s)
|
8.5 ± 2.0
|
9.5 ± 2.1
|
8.4 ± 1.8
|
0.034
|
E/A
|
1.1 ± 0.4
|
1.1 ± 0.4
|
1.0 ± 0.4
|
0.013
|
Peak E velocity (cm/s)
|
72 ± 20
|
73 ± 24
|
70 ± 18
|
0.425
|
Peak A velocity (cm/s)
|
72 ± 18
|
66 ± 16
|
74 ± 17
|
0.004
|
E/e’
|
7.9 ± 2.1
|
6.8 ± 1.7
|
8.3 ± 2.3
|
0.009
|
Tricuspid regurgitation (m/s)
|
2.3 ± 0.4
|
2.2 ± 0.4
|
2.3 ± 0.3
|
0.416
|
Data are presented as means ± standard deviation.
Reference values: LVEF >=45%, LAVI <34 ml/m2, e’ lateral >10 cm/s, e’ septal >7 cm/s, E/A 0.8 – 2.5, E/e’ 8 – 14, TR 2.0 – 2.8 m/s.
Although still in the normal range, significant differences in the diastolic function parameters were found between the lowest and highest PF quartiles. As shown in table 2, a reduced LAVI and E/e’ was found in the lowest PF quartile (p=0.02, p<0.01, respectively); and a reduced e’ lateral, e’ septal, and E/A in the highest PF quartile (p<0.01, p=0.03, p=0.01, respectively); and an increased peak A velocity in the highest PF quartile (p<0.01). Peak E velocity and TR did not differ significantly between the two extreme PF volume quartiles. Together, these differences reflect a diminished, although still normal, diastolic cardiac function in the highest PF quartile compared to the lowest PF quartile.
Association of PF with diastolic function
In the total sample (n=254), significant associations for all four diastolic parameters with the PF volume were found after adjusting for BMI, age, and sex. These data are depicted in Table 3. Analyses of the interactions with BMI, age, and sex, did not improve the model. In addition, in the extreme quartiles of PF volumes (n=130) a significantly negative association between high PF and LAVI, high PF and e’ lateral, and high PF and TR, were found after adjusting for BMI, age, and sex. However, the difference in the mobility of the septal wall between the extreme quartiles of PF volume and between E/e’ the extreme quartile of PF volume were no longer evident after the model was adjusted for these factors. These regression data are depicted in Table 4.
Table 3. Multivariable linear regression analysis in the total population exploring associations between PF and parameters of diastolic cardiac function.
|
Unadjusted regression coefficient
(95% CI)
|
p-value
|
Adjusted regression coefficient *
(95% CI)
|
p-value
|
Left atrial volume index (ml/m2)
|
-0.24 (-1.79; 1.32)
|
0.764
|
-2.05 (-3.92; -0.19)
|
0.001
|
e’ septal (cm/s)
|
-0.03 (-0.52; 0.47)
|
0.917
|
-0.13 (-0.68; 0.43)
|
0.020
|
e’ lateral (cm/s)
|
-0.21 (-0.84; 0.41)
|
0.496
|
-0.02 (-0.71; 0.67)
|
<0.001
|
E/e’
|
7.45 (6.49; 8.42)
|
0.335
|
0.16 (-0.42; 0.74)
|
0.003
|
Tricuspid regurgitation (m/s)
|
0.04 (-0.04; 0.12)
|
0.356
|
-0.02 (-0.12; 0.07)
|
0.001
|
Abbreviations: CI – confidence interval.
* Adjusted for body mass index, age, and sex
Table 4. Multivariable linear regression analysis in the extreme PF quartiles (0=low, 1=high) exploring associations between PF and parameters of diastolic cardiac function.
|
Unadjusted regression coefficient
(95% CI)
|
p-value
|
Adjusted regression coefficient *
(95% CI)
|
p-value
|
Left atrial volume index (ml/m2)
|
-4.13 (-7.47; -0.80)
|
0.015
|
-7.85 (-12.13; -3.56)
|
0.001
|
e’ septal (cm/s)
|
-1.17 (-2.25; -0.10)
|
0.034
|
-0.96 (-2.28; 0.36)
|
0.088
|
e’ lateral (cm/s)
|
-1.97 (-3.33; -0.60)
|
0.005
|
-1.39 (-3.13; 0.34)
|
0.020
|
E/e’
|
1.52 (0.40; 2.64)
|
0.009
|
1.33 (-0.11; 2.77)
|
0.118
|
Tricuspid regurgitation (m/s)
|
0.06 (-0.09; 0.22)
|
0.416
|
0.01 (-0.18; 0.20)
|
0.004
|
Abbreviations: CI – confidence interval.
* Adjusted for body mass index, age, and sex
Distribution and determinants of the different components of the PF volume
In 10% of the total sample (n=24), the EAT volume was studied by manually dividing the PF into the different CAT and EAT volumes. This random selection of 6 patients per PF quartile was made since the manual subdivision of the PF is extremely laborious, and to ascertain that the sample reflects the entire population. The data showed that with an increasing PF, no similar increase in the relative volume of EAT and CAT can be expected, as the relationship with the relative amount of EAT and CAT is lacking (p>0.7). These data are illustrated in Figure 5.
Fig. 5. No relation of PF to its CAT and EAT component. The amount of CAT (A) and EAT (B) are not related to PF. Although EAT and CAT volume show a wide variation, they are linearly associated to each other (C), indicating that both increase with an increase of PF.
To gain further insight into whether EAT is the major culprit in hampering diastolic function as suggested because of its anatomic proximity to the myocardium, separate correlations of EAT were made with the different diastolic parameters. Despite the small number, a direct correlation of the percentage of EAT and e’ lateral was found. There was no correlation with EAT and the other diastolic function parameters (Figure S4 in the Supplementary Appendix).