Introduction

Left atrial (LA) enlargement diagnosed by electrocardiography is an early sign of hypertensive heart disease.1 LA enlargement is also predictive of future ischemic stroke, atrial fibrillation and cardiac death.2, 3, 4 The factors leading to atrial enlargement in the absence of mitral valve disease are not well established. Echocardiographically determined LA size in hypertension has been related to systolic blood pressure (BP) and left ventricular (LV) hypertrophy in older patients with isolated systolic hypertension.5 Furthermore, echocardiographic studies in men with mild-to-moderate hypertension have suggested age, obesity and race as other important risk factors of LA enlargement.6

The metabolic syndrome consists of a cluster of atherosclerotic risk factors, including obesity, hypertension, insulin resistance and dyslipidemia. Many atherosclerotic risk factors are also implicated in the pathogenesis of atrial fibrillation.7, 8 A recent study in Japan has shown the relation between metabolic syndrome and the development of atrial fibrillation.9

Although obesity and insulin resistance play an important role in the pathogenesis of metabolic syndrome, the precise relation between insulin resistance and obesity to LA size has not been assessed in Japanese hypertensive patients.

Accordingly, the aim of this study was to evaluate the association of LA size with insulin resistance and obesity in nondiabetic and never-treated Japanese patients with essential hypertension. In addition, we examined the sex-specific relation of insulin resistance and echocardiographic indices of LA and LV structures.

Methods

Study population

The study population included 106 nondiabetic patients with essential hypertension (57 men, mean age: 60±12 years old; and 49 women, mean age: 62±10 years old) and 40 normotensive control (NC) individuals (21 men, 19 women, mean age: 56±14 years old). They had normal findings on chemical screening battery and were nondiabetic by the criteria of the American Diabetes Association.10 All study patients participated in this study after giving informed consent. The study was carried out in accordance with the Declaration of Helsinki (1989) of the World Medical Association. To exclude the presence of secondary forms of hypertension, atrial fibrillation and valvular heart disease, all patients underwent a complete medical history, physical examinations, echocardiographic examinations and appropriate laboratory evaluation.11

Physical examinations

Physical examinations in hypertensive patients were supervised by the nursing staff. Weight and height were measured while the subjects were fasting overnight and wearing only underwear. Body mass index (BMI) was calculated as weight (kg) divided by height (m)2. BP was measured in triplicate by a single physician who was an expert in the evaluation of hypertension, with an appropriate arm cuff and a mercury sphygmomanometer after a 5 min rest in the sitting position. The arithmetic mean of the last two measurements was calculated. Korotkoff phase V was taken for diastolic BP. Hypertension was defined as systolic BP 140 mm Hg and/or diastolic BP 90 mm Hg.

Biochemical measurements

In the morning, after an overnight fast, venous blood was sampled for the measurement of plasma concentrations of glucose and insulin, and for serum concentrations of low-density lipoprotein cholesterol, high-density lipoprotein cholesterol and triglycerides. Plasma glucose was immediately determined by the glucose oxidase method. Plasma insulin was determined in duplicate by high specific and sensitive immunoradiometric assay (Abbott Japan, Tokyo, Japan; intra-assay coefficient of variation 1.6%, interassay coefficient of variation 2.2%). Serum concentrations of low-density lipoprotein cholesterol, high-density lipoprotein cholesterol and triglycerides were assessed by standard enzymatic methods.

Insulin resistance was assessed from fasting immunoreactive insulin (FIRI) and fasting plasma glucose (FPG) and from the earlier validated homeostasis model assessment (HOMA);12 thus, HOMA value=FIRI (pmol/l) × FPG (mmol/l)/161.

Echocardiographic measurements

Two-dimensionally guided M-mode echocardiography was carried out by standard methods, as outlined earlier,11 using an SSD-6500 echocardiograph with a 3.5 MHz transducer (Aloka Inc., Tokyo, Japan). Echocardiographic examinations were carried out and interpreted by the same cardiologist, who was unaware of the patients' data. LA dimension, LV internal dimension (LVID), interventricular septal thickness and posterior wall thickness (PWT) were measured at end diastole and end systole, according to the American Society of Echocardiography Guidelines.13 LV mass was calculated according to a necropsy-validated formula.14 LV mass was also indexed by body surface area. Relative wall thickness (RWT) was measured as follows: RWT=2 × (PWTd/LVIDd), where d is end diastole. Percent fractional shortening was calculated as (LVIDd−LVIDs)/LVIDd × 100, and was used as an indicator of LV systolic function, where d and s are end-diastole and end-systole, respectively. End-diastolic and end-systolic LV volumes were calculated by the Teichholz method15 using linear measurements at diastole and systole, as has been validated by invasive and Doppler reference standards. Stroke volume (SV) was calculated as (end-diastolic LV volume–end-systolic LV volume). The ratio of SV to pulse pressure (PP) was used as an indirect measure of aortic compliance.16 Peak early transmitral flow velocity (E) and peak late transmitral flow velocity (A) were measured at the tips of mitral leaflets. The E/A ratio was calculated as an indicator of LV diastolic function. Two-dimensional and color Doppler imaging were also carried out to rule out valvular heart disease and motion abnormality.

Subgroups analysis

On the basis of the relationship between RWT and LV mass index, 106 hypertensive patients were then divided into four different groups. The partition values of 0.44 for RWT and 108 g m−2 (male) or 104 g m−2 (female) for LV mass index, the mean+2 s.d. value of NC individuals, were used. The groups consisted of hypertensive patients with normal RWT and LV mass index (normal geometry; n=43, 41%); patients with concentric remodeling (n=10, 9%); patients with concentric hypertrophy (n=19, 18%); and patients with eccentric hypertrophy (n=34, 32%).

Statistical analysis

All values are expressed as mean±s.d. One-way analysis of variance was used to evaluate the difference among groups, with Scheffe's correction for multiple comparisons. Correlation coefficients were calculated according to Pearson's method. A multiple regression analysis was also carried out to select appropriate independent variables producing the highest partial correlation with LA size in hypertensive patients. Probability values <0.05 were considered statistically significant in all analysis.

Results

Demographic and clinical characteristics

The demographic and clinical characteristics of male hypertensive patients, female hypertensive patients and NC individuals are presented in Table 1. There were no significant differences in age and pulse rate. BMI in NC individuals was significantly smaller than that in male and female hypertensive patients. However, there were no significant differences in BMI between male and female hypertensive patients. Office systolic BP, diastolic BP and PP were significantly higher in hypertensive patients than those in NC individuals. Similarly, there were no significant differences in office systolic BP, diastolic BP and PP between male and female hypertensive patients.

Table 1 Patients characteristics
Full size table

Biochemical and echocardiographic characteristics

Biochemical and echocardiographic characteristics of the three groups are presented in Table 2. FIRI, HOMA value and low-density lipoprotein cholesterol in hypertensive patients were significantly higher than those in NC individuals. Furthermore, low-density lipoprotein cholesterol in male hypertensive patients was higher than that in female hypertensive patients. High-density lipoprotein cholesterol in male hypertensive patients was lower than that in NC individuals and in female hypertensive patients. There was no significant difference in triglycerides among the three groups.

Table 2 Biochemical and echocardiographic characteristics in normotensive control individuals, male and female hypertensive patients
Full size table

The LA dimension in hypertensive patients was significantly larger than that in NC individuals. The E/A ratio was also significantly smaller in hypertensive patients than in NC individuals. The LV mass, LV mass index and RWT were largest in male hypertensive patients, followed by those in female hypertensive patients and in NC individuals. Although there was no significant difference in percent FS among the three groups, the SV/PP ratio in hypertensive patients was higher than that in NC individuals.

Relations of insulin resistance, demographic factors to LA size or LV geometry

As shown in Table 3, LA size was significantly related to BMI, HOMA value and LV mass in hypertensive patients. Table 4 shows the results of multiple regression analyses. Independent determinants of LA size in hypertensive patients were BMI and HOMA value.

Table 3 Simple correlation of left atrial size with demographic and echocardiographic variables in hypertensive patients
Full size table
Table 4 Multiple regression analysis of factors relevant to left atrial size in hypertensive patients
Full size table

As indicated in Tables 5 and 6, LV mass and RWT were significantly related to the HOMA value in male hypertensive patients. LV mass was related to BMI in female hypertensive patients. In both sexes, the LA size was significantly related to BMI and the HOMA value. Figure 1 shows the comparison of HOMA values in hypertensive patients with normal geometry (1.35±0.74), concentric remodeling (1.79±0.82), concentric hypertrophy (2.30±1.13) and eccentric hypertrophy (1.72±0.86). Hypertensive patients with concentric hypertrophy showed the highest HOMA value. Figure 2 shows the comparison of LA dimension in hypertensive patients with normal geometry (3.43±0.46 cm), concentric remodeling (3.34±0.47 cm), concentric hypertrophy (3.55±0.36 cm) and eccentric hypertrophy (3.62±0.56 cm). There were no significant correlations between LA size and LV geometry in hypertensive patients.

Table 5 Simple correlation of HOMA value with left atrial size, left ventricular mass and relative wall thickness in male and female hypertensive patients
Full size table
Table 6 Simple correlation of body mass index with left atrial size, left ventricular mass and relative wall thickness in male and female hypertensive patients
Full size table
Figure 1
figure 1

The homeostasis model assessment value associated with different left ventricular patterns in 106 patients with essential hypertension. Values are expressed as the mean±s.d.

Full size image
Figure 2
figure 2

Left atrial (LA) dimension associated with different left ventricular patterns in 106 patients with essential hypertension. Values are expressed as the mean±s.d.

Full size image

Discussion

It is widely acknowledged that the metabolic syndrome is a premier risk factor for cardiovascular disease,17, 18 and this association is partly mediated by its effect on the cardiac structure.19, 20 Consequently, we have examined the relations of insulin resistance (assessed by the HOMA value) and obesity to LV structure.21 This study added to our earlier study by identifying the significant and independent relations of insulin resistance and obesity to echocardiograhically determined LA size in nondiabetic Japanese patients with essential hypertension.

Frohlich et al.1 have suggested that the electrocardiographic evidence of LA abnormality is an early sign of hypertensive heart disease. On the other hand, insulin resistance and obesity, integral components of metabolic syndrome, are closely associated with the development of hypertension.22, 23 However, few echocardiograhic studies have assessed the relations of insulin resistance to LA size in hypertensive patients.24, 25 It is intriguing that in this study, the relation of LV mass to insulin resistance varied between the two sexes, but the LA size was related to insulin resistance in both male and female hypertensive patients. In addition, there was significant association between the LV geometric pattern and the degree of insulin resistance, but no significant correlation was present between LV geometric pattern and LA size. A possible explanation is that LA size may be influenced by clinical factors independent of their impact on LV mass and LV geometry.26

Obesity has earlier been suggested to mediate the relation between LA size and LV mass in male hypertensive patients.6 However, this study modified this conclusion by finding an independent association of BMI with LA enlargement. Furthermore, the inclusion of insulin resistance in multivariate models eliminated the correlation between LA size and LV mass. The different findings may be explained by major differences in patient characteristics, including the presence of women and less obesity. The Losartan Intervention for End Point Reduction in Hypertension (LIFE) study has shown that LA enlargement was particularly common in women.26 The mechanism by which obesity leads to LA enlargement is still under investigation. A relation to hemodynamic changes seen in obese patients, including increased intravascular volume and larger SV and cardiac output, has been suggested.27

We found in this study that an increased HOMA value and increased BMI were strongly associated with echocardiographic LA enlargement in hypertensive patients. Hypertension and obesity, integral components of the metabolic syndrome can cause atrial stretch and dilatation, resulting in a structural remodeling predisposing to atrial fibrillation.28, 29 LA enlargement is a risk for atrial fibrillation.26 Furthermore, it has been reported that metabolic syndrome is associated with LA enlargement in patients with nonvalvular atrial fibrillation.30 Another possible mechanism by which the metabolic syndrome may predispose to atrial fibrillation is inflammation and oxidative stress. Although not measured in our study, C-reactive protein and oxidants were elevated in patients with atrial fibrillation.31, 32 Insulin resistance and obesity are associated with a systemic chronic inflammatory response characterized by altered cytokine production and activation of the inflammatory signaling pathway.33, 34

The meta-analysis, based on 11 randomized controlled trials, indicates that both angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers are effective at preventing the development of atrial fibrillation.35 In hypertensive patients treated by Losartan, a reduction in the development of atrial fibrillation was evaluated.36 It is widely acknowledged that angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers have a favorable effect on insulin resistance. Therefore, these considerations provide further evidence that insulin resistance is an etiological factor for atrial fibrillation.

The influence of LV diastolic function on the left atrium has been investigated in hypertensive patients.37, 38 In our earlier study using cardiac MRI,39 we reported that LV hypertrophy was an independent determinant of LA enlargement in hypertensive patients. Theoretically, LA enlargement in hypertensive patients with LV hypertrophy can be secondary to the changes in LV diastolic function. Furthermore, we have shown that LV diastolic function was impaired even in hypertensive patients without LV hypertrophy.40 In this study, although most hypertensive patients had an impaired LV diastolic function as evaluated by the E/A ratio (abnormal relaxation pattern), the relation between LA enlargement and the E/A ratio was less striking, and the Doppler index was not identified as having an independent association with LA size in multivariate analyses.

A relation between systolic BP and LA size has been reported earlier in the Framingham Heart Study.41 Unfortunately, in this study, office systolic BP and PP were not identified as independent determinants of LA size. However, it cannot be excluded that LA enlargement may reflect, at least in part, BP elevation. Long-term average BP was more strictly related to the prevalence of LA enlargement than was in-office BP.26, 41 Therefore, long-term BP control is important for the preventions of LA enlargement and the development of atrial fibrillation.

In conclusion, there is increasing evidence of a link between metabolic syndrome and the development of atrial fibrillation, although the independent effect of metabolic syndrome on LA size is still undetermined. This study indicates that LA size is influenced by insulin resistance and obesity, integral components of the metabolic syndrome, independently of LV hypertrophy, LV geometry and LV diastolic function in nondiabetic Japanese hypertensive patients.