Comparison of baseline characteristics and chief complaints associated with hyperthyroidism (Table 1)
Table 1.. Baseline characteristics of the study population
|
AF (n=40)
|
No AF (n=96)
|
p
|
Age (years)
|
59 ± 15
|
49 ± 14
|
<0.001
|
male gender
|
18 (45%)
|
32 (33%)
|
0.242
|
SBP (mmHg)
|
119 ± 16
|
121 ± 15
|
0.477
|
DBP
|
70 ± 11
|
73 ± 10
|
0.094
|
Heart rate (bpm)
|
84 ± 19
|
84 ± 20
|
0.958
|
Body mass index (kg/m2)
|
23.7 ± 4.0
|
23.5 ± 3.6
|
0.815
|
DM
|
6 (15%)
|
9 (9%)
|
0.374
|
Hypertension
|
6 (15%)
|
23 (24%)
|
0.358
|
Alcohol
|
12 (30%)
|
16 (17%)
|
0.103
|
smoking
|
8 (20%)
|
13 (14%)
|
0.435
|
CVA
|
6 (15%)
|
2 (2%)
|
0.008
|
Chief complaints
|
|
|
0.120
|
palpitation
|
18 (45%)
|
23 (24%)
|
|
Chest pain, DOE
|
2 (5%)
|
20 (21%)
|
|
dizziness
|
3 (8%)
|
7 (7%)
|
|
weakness
|
6 (15%)
|
17 (18%)
|
|
Presence of palpitation
|
19 (48%)
|
26 (27%)
|
0.028
|
SBP: systolic blood pressure, DBP: diastolic BP, DM: diabetes mellitus, CVA: cerebrovascular attack, DOE: dyspnea on exertion
AF was diagnosed in 40 (29%) of 136 patients, and among these patients, 23 (58%) were diagnosed with PAF. The most common symptom of hyperthyroidism was palpitations.
Patients with AF were compared to those without (59 ± 15 vs. 49 ± 14 years, p<0.001), and cerebrovascular attack (CVA) was more prevalent in patients with AF than in those without (15% vs. 2%, p=0.008). Although palpitations were the most common symptom in both the groups, palpitations were more common in patients with AF than in those without (48% vs. 27%, p=0.028).
Echocardiographic and laboratory findings of the study population (Table 2)
Table 2. Echocardiographic and laboratory parameters of study population
|
AF (n=40)
|
No AF (n=96)
|
p
|
LAVI (ml/m2)
|
36.6 ± 14.9
|
26.0 ± 7.9
|
<0.001
|
LV EDD (mm)
|
49.3 ± 5.3
|
48.5 ± 4.7
|
0.381
|
LV ESD
|
31.9 ± 6.8
|
30.6 ± 4.8
|
0.218
|
LVMI (g/m2)
|
98.6 ± 23.0
|
86.4 ± 18.7
|
0.002
|
LV EF (%)
|
61.9 ± 5.9
|
66.7 ± 7.1
|
<0.001
|
E (cm/s)
|
82.0 ± 24.0
|
80.9 ± 21.1
|
0.785
|
A (cm/s)
|
68.2 ± 20.3
|
73.1 ± 20.1
|
0.298
|
E/A ratio
|
1.09 ± 0.40
|
1.19 ± 0.48
|
0.380
|
DT (ms)
|
182.9 ± 48.1
|
190.5 ± 37.0
|
0.331
|
E’ (cm/s)
|
9.0 ± 3.1
|
9.2 ± 2.7
|
0.639
|
A’ (cm/s)
|
8.9 ± 1.5
|
9.5 ± 2.3
|
0.289
|
E’/A’
|
0.88 ± 0.34
|
1.04 ± 0.43
|
0.105
|
E/E’
|
10.1 ± 3.8
|
9.3 ± 3.0
|
0.220
|
S’ (cm/s)
|
7.5 ± 1.8
|
9.1 ± 2.5
|
<0.001
|
RVSP
|
35.4 ± 9.6
|
33.0 ± 9.2
|
0.190
|
TSH (uIU/mL)
|
0.013 ± 0.013
|
0.010 ± 0.014
|
0.188
|
FT4 (ng/dL)
|
3.59 ± 2.86
|
4.00 ± 2.61
|
0.454
|
T3 (ng/dL)
|
308.0 ± 215.0
|
343.3 ± 207.9
|
0.378
|
BNP (pg/mL)
|
191.6 ± 138.1
|
115.1 ± 150.6
|
0.060
|
CK-MB (ng/mL)
|
2.63 ± 3.45
|
2.31 ± 5.79
|
0.813
|
TnI (ng/mL)
|
1.60 ± 6.65
|
0.09 ± 0.29
|
0.277
|
LAVI: left atrial volume index, LV: left ventricular, EDD: end-diastolic dimension, ESD: end-systolic dimension, LVMI: LV mass index, EF: ejection fraction, RVSP: right ventricular systolic pressure, TSH: thyroid stimulating hormone, FT4: free T 4, BNP: brain natriuretic peptide, CK-MB: creatinine kinase- musclebrain, TnI: troponin I.
The LAVI was significantly larger (37 ± 15 mL/m2 vs. 26 ± 8 mL/m2, p<0.001) in patients with AF than in those without. Although within the normal range, the LVMI was larger (99 ± 23 g/m2 vs. 86 ± 19 g/m2, p=0.002) and the LVEF was lower (62 ± 6% vs. 67 ± 7%, p<0.001) in patients with AF. Other diastolic parameters, including E/E', did not differ significantly between the two groups, but S’ velocity was lower (7.5 ± 1.8 cm/s vs. 9.1 ± 2.5 cm/s, p<0.001) in patients with AF. There was no significant difference between the two groups for blood test results of thyroid function and cardiac biomarkers, but the brain natriuretic peptide level was slightly higher in patients with AF with borderline significance than in those without (192 ± 138 pg/mL vs. 115 ± 151 pg/mL, p=0.060).
Independent predictors of atrial fibrillation in patients with hyperthyroidism (Table 3)
Table 3. Independent predictors for occurrence of atrial fibrillation in patients with hyperthyroidism
|
OR
|
95% CI
|
P
|
Univariate
|
|
|
|
age
|
1.052
|
1.022-1.084
|
0.001
|
CVA
|
8.294
|
1.597-43.087
|
0.012
|
palpitation
|
2.436
|
1.132-5.243
|
0.023
|
LAVI
|
1.101
|
1.052-1.153
|
<0.001
|
LVMI
|
1.030
|
1.010-1.051
|
0.003
|
EF
|
0.902
|
0.848-0.960
|
0.001
|
S’
|
0.699
|
0.570-0.857
|
0.001
|
Multivariate
|
|
|
|
age
|
1.017
|
0.979-1.057
|
0.390
|
CVA
|
4.087
|
0.493-33.872
|
0.192
|
palpitation
|
5.636
|
1.987-15.986
|
0.001
|
LAVI
|
1.092
|
1.030-1.159
|
0.003
|
LVMI
|
1.010
|
0.982-1.038
|
0.480
|
LV EF
|
0.967
|
0.890-1.051
|
0.428
|
LV S’
|
0.834
|
0.649-1.072
|
0.157
|
CVA: cerebrovascular attack, LAVI: left atrial volume index, LVMI: left ventricular mass index, EF: ejection graction. S’: S’ velocity.
Younger age, CVA, presence of palpitations, larger LAVI, higher LVMI, and lower EF and S’ velocity were found to be associated with AF in the univariate analysis. Among them, presence of palpitations (odds ratio [OR]: 5.636, 95% confidence interval [CI]: 1.987–15.986, p=0.001) and larger LAVI (OR: 1.092, 95% CI: 1.030–1.159, p=0.003) were found to be independently associated with the occurrence of AF in patients with hyperthyroidism.
Left atrial expansion index, ejection fraction, and strain analysis (Table 4)
Table 4. LA EI, EF and strain, strain rate and other parameters of the study population
|
AF (n=40)
|
No AF (n=96)
|
p
|
LA EI (%)
|
74.5 ± 46.9
|
126.6 ± 69.4
|
0.001
|
LA EF (%)
|
49.8 ± 16.0
|
63.3 ± 10.6
|
<0.001
|
PALS
|
19.3 ± 9.9
|
27.6 ± 9.5
|
<0.001
|
PACS*
|
9.0 ± 5.4
|
12.8 ± 4.1
|
0.002
|
LA E SR
|
-1.07 ± 0.61
|
-1.31 ± 0.59
|
0.085
|
LA A SR*
|
-1.10 ± 0.59
|
-1.67 ± 0.55
|
<0.001
|
LA S SR
|
0.95 ± 0.46
|
1.37 ± 0.45
|
<0.001
|
Correlation between LAVI and LA strain and strain rate
|
Variables
|
Pearson correlation
|
p
|
LAEI (%)
|
-0.338
|
0.002
|
LA EF (%)
|
-0.481
|
<0.001
|
PALS
|
-0.320
|
0.004
|
PACS
|
-0.416
|
<0.001
|
LA A sr
|
-0.407
|
<0.001
|
LAS sr
|
-0.367
|
0.001
|
LA: left atrial, R:reservoir function, LA EF: LA ejection fraction. LAEI : LA expansion index, PALS: peak atrial longitudinal strain, PACS: peak atrial contraction strain, LA E, A and S sr: LA E, A and S strain rate, *: in patients with PAF (sinus rhythm during echocardiography)
Table 4 shows the additional LA volumetric and strain analysis. The LA EI (75 ± 47% vs. 127 ± 69%, p<0.001) and LA EF (75 ± 47% vs. 127 ± 69%, p<0.001) were significantly decreased in patients with AF than in those without. Figure 1 shows the ROC curve for LAVI (cut-off value≥24 mL/m2, area under the curve [AUC]: 0.769, SE: 0.055, p<0.001), LA EI (cut-off value≤81%, AUC: 0.224, SE: 0.056, p<0.001), and LA EF (cut-off value≤56%, AUC: 0.224, SE: 0.058, p<0.001). In the strain analysis, both PALS and PACS were decreased in patients with AF. The strain rate analysis showed that LA A sr and S sr were decreased in patients with AF. The LA EI, EF, PALS, PACS, A sr, and S sr all showed negative correlations with the LAVI. Figure 2 shows each of these correlations.