CiteSeerX — Adiponectin Level and Left Ventricular Hypertrophy in

(1)

and Hideaki Toyoshima

Keiko Wada, Kaichiro Sugiura, Seiko Takefuji, Yo Hotta, Takahisa Kondo, Toyoaki Murohara Hirotsugu Mitsuhashi, Hiroshi Yatsuya, Koji Tamakoshi, Kunihiro Matsushita, Rei Otsuka,

Adiponectin Level and Left Ventricular Hypertrophy in Japanese Men

Print ISSN: 0194-911X. Online ISSN: 1524-4563

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Hypertension

doi: 10.1161/HYPERTENSIONAHA.106.079509

2007;49:1448-1454; originally published online April 9, 2007;

Hypertension.

http://hyper.ahajournals.org/content/49/6/1448

World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://hyper.ahajournals.org//subscriptions/

is online at:

Hypertension Information about subscribing to

Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at:

Reprints:

document.

Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information about Office. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Hypertension

in

Requests for permissions to reproduce figures, tables, or portions of articles originally published Permissions:

by guest on February 27, 2014 http://hyper.ahajournals.org/

Downloaded from http://hyper.ahajournals.org/ by guest on February 27, 2014 Downloaded from

(2)

Adiponectin Level and Left Ventricular Hypertrophy in Japanese Men

Hirotsugu Mitsuhashi, Hiroshi Yatsuya, Koji Tamakoshi, Kunihiro Matsushita, Rei Otsuka, Keiko Wada, Kaichiro Sugiura, Seiko Takefuji, Yo Hotta, Takahisa Kondo,

Toyoaki Murohara, Hideaki Toyoshima

Abstract—A recent study has demonstrated that adiponectin inhibited hypertrophic signaling in the myocardium of mice, implying that a decrease in the blood adiponectin level could cause cardiac muscle hypertrophy. We hypothesized that a relationship might exist between the serum adiponectin level and electrocardiographically diagnosed left ventricular hypertrophy (ECG-LVH), and we examined this hypothesis by epidemiological study of 2839 Japanese male workers who were not taking medications for hypertension. ECG-LVH was defined as meeting Sokolow–Lyon voltage criteria and/or Cornell voltage-duration product. The subjects were categorized by tertiles of serum adiponectin level, and a multivariate logistic regression analysis was conducted relating left ventricular hypertrophy to adiponectin tertiles adjusting for potential confounding factors. Prevalence of ECG-LVH in the studied sample was 16.7%. Adiponectin ranged from 1.0 to 5.0␮g/mL in the lowest category and from 7.4 to 30.6 ␮g/mL in the highest. Compared with subjects in the highest adiponectin category, those in the lowest one had a significantly higher prevalence of ECG-LVH independent of age, body mass index, and systolic blood pressure with an odds ratio of 1.50 and a 95%

CI of 1.16 to 1.94. Further adjustment for high-density lipoprotein cholesterol, triglyceride, and insulin resistance did not change the association (odds ratio: 1.68; 95% CI: 1.28 to 2.21; P⬍0.001). Similar results were obtained when different criteria for ECG-LVH were used or when subjects were stratified by blood pressure or body mass index. Adiponectin concentration was inversely and independently associated with ECG-LVH in Japanese men.

(Hypertension. 2007;49:1448-1454.)

Key Words: adiponectin 䡲 left ventricular hypertrophy 䡲 electrocardiography 䡲 epidemiologic study 䡲 Japanese

A

diponectin is a serum protein of 30 kDa, which is a member of the complement factor C1q family and is produced and secreted by adipose tissue.¹ Its associations with cardiac risk factors, such as metabolic syndrome, type 2 diabetes, and hypertension, have been described.^{2– 4} One recent experimental study demonstrated that adiponectin inhibited hypertrophic signaling in the myocardium by inhibiting extracellular signal regulated kinase and activating adenosine monophosphate-activated protein kinase in mice.⁵

Left ventricular hypertrophy (LVH) is characterized by an increase in chamber mass produced largely by an increase in the size of cardiomyocytes.⁶ The determinants of LVH include age, elevated blood pressure (BP), obesity, and insulin resistance.⁷ LVH diagnosed by ECG, as well as echocardiographically diagnosed LVH, has been known to be associated with increased risk of cardiovascular morbidity and mortality.^{8 –13} The prevalence of ECG-LVH in Japan

increased from 15.3% in 1990 to 19.7% in 2000,^14,15which warrants an investigation of LVH in relation to its determinants aiming at reducing cardiovascular morbidity and mortality.

Recently, a case–reference study enrolling a small sample mainly including hypertensive patients reported the inverse association between echocardiographically determined left ventricular mass index and blood adiponectin level.¹⁶How- ever, hypertensive states may have confounded the association, because hypertension could cause both LVH and reduction of adiponectin concentration via arteriosclerosis. In addition, some antihypertensive agents are known to reduce left ventricular wall thickness and also to increase adiponectin levels, further complicating the association.^17–19 To gain epidemiological insight into the pathogenesis of LVH in relation to adiponectin, it is necessary to examine the association in healthy individuals not taking medica- tion for hypertension that could influence LVH and adi-

Received August 31, 2006; first decision September 25, 2006; revision accepted March 18, 2007.

From the Department of Public Health/Health Information Dynamics (H.M., H.Y., K.T., K.M., R.O., K.W., K.S., S.T., Y.H., H.T.), Nagoya University Graduate School of Medicine, Nagoya, Japan; the Department of Cardiology (H.M., K.M., K.S., T.K., T.M.), Nagoya University Graduate School of Medicine, Nagoya, Japan; and the Department of Endocrinology (S.T., Y.H.), Nagoya University Graduate School of Medicine, Nagoya, Japan.

Correspondence to Hideaki Toyoshima, Department of Public Health/Health Information Dynamics, Nagoya University Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466-8550, Japan. E-mail hirot@med.nagoya-u.ac.jp

Hypertension is available at http://www.hypertensionaha.org DOI: 10.1161/HYPERTENSIONAHA.106.079509 1448

(3)

ponectin concentration. Therefore, we conducted a large- scale epidemiological study in Japanese male workers, examining the association between blood adiponectin level and LVH. We also investigated the association while stratifying subjects by BP level that might confound the association.

Methods

Study Population

This study uses baseline data collected in 2002 for a workers’ cohort study on cardiovascular diseases in Aichi, Japan. The present analysis included 3168 men aged 35 to 66 years old with available information on weight, height, BP, and ECG, as well as adiponectin and other biomarker concentrations. Written consent was obtained from each person, and the study protocol was approved by the ethics review committee of the Nagoya University Graduate School of Medicine. We excluded 259 subjects who had been on medical treatment for hypertension from the initial 3168 subjects. Because the diagnosis of LVH by ECG is not reliable in subjects with myocardial infarction or complete bundle branch block, 5 subjects with ECG findings relevant for a history of myocardial infarction and 65 subjects who had ECG findings of complete bundle branch block were excluded. Finally, 2839 men were included in the present analysis.

ECG

Standard 12-lead ECG was recorded at a paper speed of 25 mm/s and 1 mV/cm of standardization. Every ECG was read independently by 2 experienced investigators blinded to the information of the health checkups according to a slightly modified 1982 revision of the Minnesota Code. QRS duration and voltage of R waves in leads aVL

(RaVL), V5(RV5), and V6(RV6), and S waves in leads V1(SV1) and V3(SV3) were measured manually by a third cardiologist. LVH in this study was defined using the following 2 criteria, which the Losartan Intervention for Endpoint Reduction in Hypertension study group used^20,21: the Cornell product obtained as Cornell voltage (RaVL⫹SV3)⫻QRS duration ⬎2440 mm⫻ms and Sokolow–Lyon voltage (SV1⫹RV5or RV6)⬎38 mm. Subjects meeting either or both of these criteria were defined as having ECG-LVH.^20,21

To examine the effect of change in the definition of ECG-LVH, 4 other ECG-LVH criteria were set as follows: (1) Minnesota Code, 3-1 (RV5or RV6⬎2.6 mV; R in leads I, II, III, or RaVF⬎2.0 mV;

or RaVL⬎1.2 mV) or 3-3 (1.5 mV⬍RIⱕ2.0 mV; SV1⫹RV5; or RV6

⬎3.5 mV)²²; (2) Sokolow–Lyon voltage ⬎38 mm; (3) Cornell

voltage (RaVL⫹SV3) ⱖ28 mm^20,21; and (4) Cornell product

⬎2440 mm䡠ms.²³

Adiponectin Concentration

Venous blood samples were withdrawn from each subject afterⱖ8 hours fasting. The samples were stored at⫺80°C until assay. Serum concentration of adiponectin was measured by an ELISA system (Otsuka Pharmaceutical Co, Ltd). Interassay coefficients of variation of adiponectin were⬍8.6%.

Clinical Features and Other Laboratory Measurements

Height and weight were measured in standing position. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters, both of which were obtained during an annual health checkup. BP was measured in sitting position by well-trained nurses afterⱖ5 minutes of rest by auscultation with an appropriate arm cuff and a mercury column manometer in 66% of the subjects analyzed and by an automatic sphygmomanometer with oscillometric method (BP-103iII; Omron-Colin Co) in 34%. Pulse pressure was calculated as systolic BP (SBP) minus diastolic BP (DBP). The following items were also determined from the stored serum: total cholesterol, triglyceride (TG), high-density lipoprotein cholesterol (HDL-chol), fasting glucose, and homeostasis model assessment of insulin resistance (HOMA-R). HOMA-R was calculated as insulin (microunits per milliliter)⫻fasting glucose (mil- limoles per liter)/22.5. Total cholesterol and TG were measured enzymatically. HDL-chol was determined by the phosphotung- state method. Fasting glucose was enzymatically measured by the hexokinase method. Insulin was measured by radioimmunoassay system.

Statistical Analysis

Because distribution of adiponectin, TG, HDL-chol, and HOMA-R were skewed, log-transformed value was used for the subsequent analyses to approximately normalize the distributions. These log- transformed continuous variables were expressed as geometric means with the 95% CIs. Other continuous variables were expressed as means with the 95% CI. Subjects were divided into 3 categories by tertiles of serum adiponectin level. Comparisons of continuous variables among 3 categories were performed by 1-way ANOVA and categorical variables by␹²test. The odds ratios (ORs) and their 95%

CIs of LVH according to adiponectin levels were calculated by multivariate logistic regression analysis adjusting for potential confounding factors, that is, age, BMI, TG, HDL-chol, HOMA-R, and either SBP or pulse pressure.^7,24The same analyses were conducted TABLE 1. Baseline Characteristics of Study Sample According to Adiponectin Tertiles

Variables

Tertiles

P*

Lowest (n⫽953) Middle (n⫽940) Highest (n⫽946) Adiponectin Ranges,␮g/mL

1.0 to 5.0 5.1 to 7.3 7.4 to 30.6

Age, y 48.1 (47.6 to 48.5) 48.1 (47.6 to 48.5) 48.5 (48.0 to 49.0) 0.293

BMI, kg/m² 24.2 (24.0 to 24.3) 23.3 (23.1 to 23.4) 22.1 (21.9 to 22.3) ⬍0.001 SBP, mm Hg 127.9 (127.3 to 129.1) 127.9 (126.9 to 128.8) 126.1 (125.2 to 127.0) 0.002 Adiponectin,␮g/mL 3.72 (3.66 to 3.78) 6.1 (6.06 to 6.14) 9.86 (9.71 to 10.02) ⬍0.001 T-chol, mmol/L 5.54 (5.49 to 5.60) 5.47 (5.41 to 5.53) 5.33 (5.28 to 5.39) ⬍0.001 TG, mmol/L 3.78 (3.65 to 3.92) 3.16 (3.05 to 3.27) 2.38 (2.30 to 2.45) ⬍0.001 HDL-chol, mmol/L 1.27 (1.25 to 1.29) 1.4 (1.38 to 1.43) 1.58 (1.55 to 1.66) ⬍0.001 FBG, mmol/L 5.42 (5.35 to 5.48) 5.31 (5.26 to 5.36) 5.27 (5.21 to 5.33) 0.001 HOMA-R 1.92 (1.83 to 2.02) 1.57 (1.49 to 1.65) 1.24 (1.18 to 1.31) ⬍0.001

LVH, n (%) 189 (19.8) 141 (15.0) 144 (15.2) 0.006

T-chol indicates total cholesterol; FBG, fasting blood glucose. All means (age, BMI, SBP, and T-chol) and geometric means (adiponectin, TG, HDL-chol, and HOMA-R) are expressed with 95% CIs. LVH is expressed as n, %.

*Differences in mean values were tested by 1-way ANOVA. Difference in the proportions were tested by␹²test.

Mitsuhashi et al Adiponectin and LVH in Japanese Men 1449

Downloaded from

(4)

in samples stratified by BP and BMI. High BP was defined as SBP of ⱖ140 mm Hg and/or DBP of ⱖ90 mm Hg. Overweight was defined as BMI of ⱖ25 kg/m². In addition, we presented and compared the prevalence of ECG-LVH among the following 4 groups: normal weight with normal BP, normal weight with high BP, overweight with normal BP, and overweight with high BP. A P⬍0.05 was considered statistically significant. All of the analyses and calculations were performed using the statistical package SPSS 12.0 J for Windows.

Results

The mean age of the present sample was 48.2 years old. BMI ranged from 15.0 to 35.8 kg/m², and the mean was 23.7 kg/m². SBP ranged from 84 to 191 mm Hg, and the mean was 127.4 mm Hg. Serum adiponectin level ranged from 1.0 to 30.6 ␮g/mL, and the mean was 6.1 ␮g/mL. Prevalence of ECG-LVH was 16.7%. Baseline characteristics of the study sample according to serum adiponectin categories are shown in Table 1. Significant differences among 3 categories were found in BMI, SBP, total cholesterol, TG, HDL-chol, fasting glucose, and HOMA-R. Prevalence of LVH in the lowest, middle, and the highest categories were 19.8%, 15.0%, and 15.2%, respectively.

Compared with the category with the highest adiponectin, in the category with the lowest adiponectin, the OR of LVH was 1.38 (95% CI: 1.09 to 1.75; P⫽0.008) in the univariable logistic analysis and 1.50 (95% CI: 1.16 to 1.94; P⫽0.002) in the multivariate logistic regression analysis adjusted for age, BMI, and SBP (Table 2). When all of the potential confounding factors for LVH, that is, age, BMI, SBP, TG, HDL-chol, and HOMA-R, were adjusted, the OR in the lowest category was 1.68 (95% CI: 1.28 to 2.21; P⬍0.001). An almost identical result was obtained when pulse pressure was adjusted instead of SBP (OR: 1.67; 95% CI: 1.27 to 2.19;

P⬍0.001). ORs calculated in these 2 analyses were higher than the OR of 10 mm Hg increments of SBP (OR⫽1.54;

P⬍0.001) in the final model. The OR in the middle category was not significantly different from unity in all of the analyses above.

Prevalence of ECG-LVH was significantly higher in the high-BP group than in normal-BP group in both BMI categories (Figure). However, there was no statistically signifi-

cant difference in the prevalence of ECG-LVH between the BMI groups.

In the category with the lowest adiponectin, multivariate- adjusted OR by which subjects had LVH was 1.65 (95% CI:

1.18 to 2.30; P⫽0.003) and 1.75 (95% CI: 1.07 to 2.86;

P⫽0.027) in the normal-BP and high-BP groups, respectively (Table 3). The same analyses were conducted when the sample was divided into 2 groups depending on BMI (Table 4). In the normal-BMI group, the OR in the lowest adiponectin category was 1.45 (95% CI: 1.07 to 1.98;

P⫽0.018). In the overweight group, the OR was 1.65 (95%

CI: 0.95 to 2.86; P⫽0.074). Changing the definition of ECG-LVH did not substantially influence the present findings (Table 5).

Discussion

Our study revealed the significant inverse and independent association of LVH determined electrocardiographically with the serum adiponectin level in Japanese men. Although

Role of HTN and obesity on ECG-LVH (n⫽2839). HTN indicates hypertension. Difference between the normal-BP and high-BP group was calculated by␹²test. Normal BP and high BP were defined as SBP⬍140 mm Hg and DBP ⬍90 mm Hg and SBP ⱖ140 mm Hg or DBP ⱖ90 mm Hg, respectively. Normal weight and overweight were defined as BMI⬍25 kg/m²and BMIⱖ25 kg/m², respectively.

TABLE 2. Association Between Adiponectin Tertiles and Prevalence of LVH by Multivariate Logistic Regression Analysis (nⴝ2839)

Adiponectin Tertiles n No. of LVH (%)

Crude Age, BMI, and SBP Adjusted Multivariate Adjusted*

OR 95% CI P OR 95% CI P OR 95% CI P

Lowest 953 189 (19.8) 1.38 1.09 to 1.75 0.008 1.50 1.16 to 1.94 0.002 1.68 1.28 to 2.21 ⬍0.001 Middle 940 141 (15.0) 0.98 0.76 to 1.27 0.89 1.00 0.77 to 1.30 0.98 1.06 0.81 to 1.38 0.69

Highest 946 144 (15.2) 1 䡠䡠䡠䡠䡠䡠 1 䡠䡠䡠䡠䡠䡠 1 䡠䡠䡠䡠䡠䡠

Age, BMI, and PP Adjusted Multivariate Adjusted†

Lowest 1.52 1.18 to 1.56 0.001 1.67 1.27 to 2.19 ⬍0.001

Middle 1.04 0.81 to 1.35 0.75 1.09 0.84 to 1.42 0.52

Highest 1 䡠䡠䡠䡠䡠䡠 1 䡠䡠䡠䡠䡠䡠

PP indicates pulse pressure.

*Adjusted for age, BMI, SBP, and log-transformed HDL-chol, TG, and HOMA-R.

†Adjusted for age, BMI, pulse pressure, log-transformed HDL-chol, TG, and HOMA-R.

(5)

echocardiographic criteria are regarded as the gold standard for diagnosis of LVH, ECG has been widely used in clinical practice and periodic health checkups because of its avail- ability and low cost and is still considered the standard method for screening of LVH.²⁵ ECG is also used in epidemiological studies enrolling large populations, and ECG-LVH has been confirmed as one of the risk factors of cardiovascular morbidity and mortality.^11,25 Assuming the low adiponectin level (lowest tertile in the present study) is a causal factor of LVH, the population-attributable risk fraction was calculated to be 0.19, that is, 19% of LVH could be prevented or regressed by interventions to increase the adiponectin level adequately.

The major finding of the present study was the significant relationship between serum adiponectin and ECG-LVH in apparently healthy individuals with normal and high BP. The association was independent of age, the degree of obesity, variability in BP, and other potential confounding factors, including insulin resistance and lipids.^{24,26 –29} This finding would be consistent with the experimental studies, which reported that adiponectin-knockout mice developed enhanced concentric cardiac hypertrophy, and injection of adiponectin to LVH mice significantly reduced LVH.⁵The present study extended the current understanding of the association observed in a small sample mainly composed of hypertensive patients.¹⁶ We have here demonstrated the association in more detail in a large sample of apparently healthy individuals and also revealed that the association was independent of age and the degree of obesity evaluated by BMI, BP or pulse pressure, lipid parameters, and index of insulin resistance.

We excluded subjects on medical treatment for hypertension and examined this association in a normal-BP subgroup in which subjects with elevated BP were also excluded to minimize the confounding effects of hypertension or medications for hypertension.^17,18 Hypertension is acknowledged to be a major determinant of LVH,³⁰and in the present study

it increased the prevalence of LVH approximately twice compared with normotensive subjects (Figure). Hypertension could cause or have an association with decreased adiponectin concentration as well.⁴Thus, the present findings suggest that adiponectin may be causally associated with LVH, although this study is cross-sectional.

On the other hand, compared with the category with the highest adiponectin, in the category with an intermediate amount of adiponectin, the OR of LVH was not statistically significantly different from unity. In an additional analysis in which the higher 2 adiponectin categories were combined, the lowest tertile was significantly associated with a higher prevalence of LVH with similar magnitude (OR: 1.63; 95%

CI: 1.30 to 2.04; P⬍0.001). In a study of Japanese adults, a clinical entity of hypoadiponectinemia (⬍4.0␮g/mL) was suggested in relation to metabolic syndrome.³¹The lowest adiponectin category in the present study included subjects whose adiponectin levels were ⬍4.0 ␮g/mL. Taken to- gether, the findings imply that a threshold adiponectin level to induce LVH may exist; thus, only very low levels of adiponectin may have adverse effects related to LVH.

Further study is warranted to examine whether such a threshold exists.

Although elucidating the mechanism by which adiponectin might prevent LVH is beyond the scope of the present study, one possibility might be that adiponectin has the ability to suppress hypertrophic remodeling of the myocardium. Adi- ponectin directly inhibited hypertrophic signaling in the myocardium by activating adenosine monophosphate-activated protein kinase, which activates eukaryotic elongation factor-2 kinase and inhibits protein synthesis of cardiac myocytes.^5,27,32

In the present study, we used ECG in the diagnosis of LVH, and ECG-LVH was defined as a finding that complied with the Sokolow–Lyon voltage and/or the Cornell product, which were reported to have significantly higher sensitivity (40% to 50%) and specificity (⬇95%) among various criteria TABLE 3. Association Between Adiponectin Tertiles and Prevalence of LVH by Multivariate Logistic

Regression Analysis According to BP (nⴝ2839)

Adiponectin Tertiles n/N of LVH (%)

SBP⬍140 and DBP ⬍90 mm Hg (n⫽2245)

n/N of LVH (%)

SBPⱖ140 or DBP ⱖ90 mm Hg (n⫽594)

OR* 95% CI P OR* 95% CI P

Lowest 117/737 (15.9) 1.65 1.18 to 2.30 0.003 64/200 (32.0) 1.75 1.07 to 2.86 0.027 Middle 100/752 (13.3) 1.19 0.86 to 1.63 0.29 45/195 (23.1) 1.03 0.63 to 1.69 0.92

Highest 96/755 (12.7) 1 䡠䡠䡠䡠䡠䡠 52/199 (26.1) 1 䡠䡠䡠䡠䡠䡠

*Adjusted for age, BMI, SBP, and log-transformed HDL-chol, TG, and HOMA-R.

TABLE 4. Association Between Adiponectin Tertiles and Prevalence of LVH by Multivariate Logistic Regression Analysis According to BMI (nⴝ2839)

Adiponectin Tertiles n/N of LVH (%)

BMI⬍25 kg/m²(n⫽2148)

n/N of LVH (%)

BMIⱖ25 kg/m²(n⫽691)

OR* 95% CI P OR* 95% CI P

Lowest 139/715 (19.4) 1.45 1.07 to 1.98 0.018 41/230 (17.8) 1.65 0.95 to 2.86 0.074 Middle 109/704 (15.5) 1.05 0.77 to 1.42 0.76 37/235 (15.7) 1.23 0.72 to 2.09 0.44

Highest 115/729 (15.8) 1 䡠䡠䡠䡠䡠䡠 33/226 (14.6) 1 䡠䡠䡠䡠䡠䡠

*Adjusted for age, BMI, SBP, log-transformed HDL-chol, TG, and HOMA-R.

Downloaded from

(6)

when echocardiography-diagnosed LVH was used as the gold standard.^21,23,33This criterion is known to be influenced by BMI, that is, the sensitivity has been reported to be lower in subjects with higher BMI because of the waning QRS voltage in such individuals. Waning of QRS voltage has been reported to occur in those subjects because the distance between the electrode and the left ventricle tends to become greater.^20,33–37 This may be because of the increment of adipose mass and clockwise cardiac rotation, which were frequently observed characteristics of overweight and obese subjects.^20,38,39 We, thus, conducted an analysis to investigate whether clockwise rotation might influence the result. OR was similar even when clockwise rotation was added to the adjustment factors, which were used in the analysis model on Table 4 in the overweight group (OR:

1.66; 95% CI: 0.95 to 2.90; P⫽0.075). However, because the mean BMI in the present sample was relatively low (23.7 kg/m²), the low sensitivity would not have caused significant distortion in the association between adiponectin and LVH. Furthermore, we enhanced the reliability of the results by obtaining a similar finding when different diagnostic criteria were used.

Nevertheless, it may still be argued that only a borderline significant association was obtained in the overweight group, although the association was significant in the normal-BMI group. As stated earlier, underestimation of the prevalence of ECG-LVH might have attenuated the significance, although its magnitude was unknown. However, the point estimate of OR, which would represent the strength of the association, was greater in the overweight group than in the normal-BMI group. Thus, we speculated that a lack of statistical power because of a limited number of overweight subjects would be a major reason behind the borderline significance. In the future, studies with echocardiography to diagnose LVH are warranted, especially in overweight and obese subjects, to confirm the present findings.

Although we have adjusted for the degree of obesity by using BMI, there may be a possibility that adjustment with other measures of obesity, such as waist circumference or waist/hip ratio, alters the present association. However, BMI is known to be a useful predictor of morbidity and mortality and is highly correlated with waist circumference in Japanese subjects.^{40 – 42}BMI, as well as waist circumference, has been reported to be significantly associated with LVH to the same degree.⁴³They also had similar correlation coefficients with the adiponectin level.⁴⁴

Because renal function has been reported to be associated with LVH and adiponectin,^45,46mild renal impairment may also be a potential confounding factor. We conducted another logistic regression analysis, further adjusting for serum creatinine level in a subsample whose creatinine level was available (n⫽2589) and found almost identical results (OR by which subjects had LVH in the lowest category against the highest one: 1.67; P⬍0.001).

We have excluded medically treated hypertensive patients and further statistically controlled the confounding effects of hypertension by adjusting for BP measured at the annual health checkup. We also conducted a stratified analysis by BP and found similar results in both strata. Because ambulatory or home BP measurements may detect other types of hypertension, such as out-of-clinic hypertension and home hypertension, which have also been associated with LVH,^30,47future studies should obtain these measurements to confirm the present findings. In conclusion, the adiponectin level was inversely associated with ECG-LVH in Japanese men independent of age, BP, obesity, lipids, and insulin resistance.

Perspectives

Adiponectin might be causally associated with LVH independent of age, BP, the degree of obesity, lipids, and insulin resistance. The findings should first be confirmed by studies with echocardiography. Further investigations in women, other ethnic groups, or other populations with different body habitus should also be conducted. Ideally, a prospective cohort study or interventional study should be conducted to examine whether increasing adiponectin levels in LVH patients may regress LVH in the future.

Sources of Funding

This work was supported by grants to H.T. (17390185), H.Y.

(17790384), and K.T. (18590594) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan, and a research grant from the Japan Atherosclerosis Prevention Fund.

Disclosures

None.

References

1. Iacobellis G, Pistilli D, Gucciardo M, Leonetti F, Miraldi F, Brancaccio G, Gallo P, di Gioia CR. Adiponectin expression in human epicardial adipose tissue in vivo is lower in patients with coronary artery disease.

Cytokine. 2005;29:251–255.

TABLE 5. Association of Adiponectin Level With LVH Defined by Several ECG Criteria (nⴝ2839)

Criteria No. of LVH (%)

Middle (n⫽940) Lowest (n⫽953)

OR 95% CI P OR 95% CI P

Minnesota Code 596 (21.0) 1.24 0.98 to 1.58 0.076 1.48 1.14 to 1.91 0.003 Sokolow–Lyon voltage 339 (11.9) 1.02 0.75 to 1.39 0.88 1.58 1.15 to 2.16 0.004

Cornell voltage 132 (4.7) 0.95 0.56 to 1.62 0.84 1.71 1.02 to 2.85 0.041

Cornell product 163 (5.7) 1.08 0.70 to 1.66 0.72 1.59 1.03 to 2.45 0.037

Sokolow–Lyon voltage or/and Cornell product* 474 (16.7) 1.06 0.81 to 1.38 0.69 1.68 1.28 to 2.21 ⬍0.001 Variables adjusted for in the analysis include age, BMI, SBP, log-transformed TG, HDL-chol and HOMA-R.

*Same as the multivariate-adjusted model in Table 2.

(7)

2. Matsuzawa Y, Funahashi T, Kihara S, Shimomura I. Adiponectin and metabolic syndrome. Arterioscler Thromb Vasc Biol. 2004;24:29 –33.

3. Matsushita K, Yatsuya H, Tamakoshi K, Wada K, Otsuka R, Takefuji S, Sugiura K, Kondo T, Murohara T, Toyoshima H. Comparison of circu- lating adiponectin and proinflammatory markers regarding their asso- ciation with metabolic syndrome in Japanese men. Arterioscler Thromb Vasc Biol. 2006;26:871– 876.

4. Iwashima Y, Katsuya T, Ishikawa K, Ouchi N, Ohishi M, Sugimoto K, Fu Y, Motone M, Yamamoto K, Matsuo A, Ohashi K, Kihara S, Funahashi T, Rakugi H, Matsuzawa Y, Ogihara T. Hypoadiponectinemia is an independent risk factor for hypertension. Hypertension. 2004;43:

1318 –1323.

5. Shibata R, Ouchi N, Ito M, Kihara S, Shiojima I, Pimentel DR, Kumada M, Sato K, Schiekofer S, Ohashi K, Funahashi T, Colucci WS, Walsh K.

Adiponectin-mediated modulation of hypertrophic signals in the heart.

Nat Med. 2004;10:1384 –1389.

6. Walsh RA. Molecular and cellular biology of the normal, hypertrophied, and failing heart. In: Fuster V, Alexander RW, O’Rourke RA, Roberts R, King III SB, Wellens HJ. Hurst’s The Heart, 10th ed. New York, NY:

McGraw-Hill; 2001:115–125.

7. Ilercil A, Devereux RB, Roman MJ, Paranicas M, O’Grady MJ, Welty TK, Robbins DC, Fabsitz RR, Howard BV, Lee ET. Relationship of impaired glucose tolerance to left ventricular structure and function: the Strong Heart Study. Am Heart J. 2001;141:992–998.

8. Thrainsdottir IS, Hardarson T, Thorgeirsson G, Sigvaldason H, Sigfusson N. Survival and trends of occurrence of left ventricular hypertrophy, gender differences, 1967–92: the Reykjavik Study. J Intern Med. 2003;

253:418 – 424.

9. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically-determined left ventricular mass in the Framingham Heart Study. N Engl J Med. 1990;322:1561–1566.

10. Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: a risk profile from the Framingham Study. Stroke. 1991;22:

312–318.

11. Kannel WB. Left ventricular hypertrophy as a risk factor: the Fra- mingham experience. J Hypertens. 1991;9(suppl):S3–S8.

12. Bots ML, Nikitin Y, Salonen JT, Elwood PC, Malyutina S, Freire de Concalves A, Sivenius J, Di Carlo A, Lagiou P, Tuomilehto J, Koudstaal PJ, Grobbee DE. Left ventricular hypertrophy and risk of fatal and non-fatal stroke. EUROSTROKE: a collaborative study among research centres in Europe. J Epidemiol Community Health. 2002;56(suppl 1):

i8 –i13.

13. Okin PM, Devereux RB, Nieminen MS, Jern S, Oikarinen L, Viitasalo M, Toivonen L, Kjeldsen SE, Julius S, Snapinn S, Dahlof B; LIFE Study Investigators. Electrocardiographic strain pattern and prediction of cardio- vascular morbidity and mortality in hypertensive patients. Hypertension.

2004;44:48 –54.

14. Nakamura K, Okamura T, Hayakawa T, Kadowaki T, Kita Y, Okayama A, Ueshima H; NIPPON DATA90 Research Group. Electrocardiogram screening for left high R-wave predicts cardiovascular death in a Japanese community-based population: NIPPON DATA90. Hypertens Res. 2006;

29:353–360.

15. Ministry of Health, Labour and Welfare. The Fifth National Survey of Cardiovascular Diseases. Tokyo, Japan: Ministry of Health, Labour and Welfare; 2002:291.

16. Hong SJ, Park CG, Seo HS, Oh DJ, Ro YM. Associations among plasma adiponectin, hypertension, left ventricular diastolic function and left ven- tricular mass index. Blood Press. 2004;13:236 –242.

17. Dahlof B, Devereux R, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Julius S, Kjeldsen S, Kristianson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H. The Losartan Intervention For Endpoint Reduction (LIFE) in Hypertension Study:

rationale, design, and methods. The LIFE Study Group. Am J Hypertens. 1997;10:705–713.

18. Okin PM, Devereux RB, Jern S, Kjeldsen SE, Julius S, Nieminen MS, Snapinn S, Harris KE, Aurup P, Edelman JM, Wedel H, Lindholm LH, Dahlof B; LIFE Study Investigators. Regression of electrocardiographic left ventricular hypertrophy during antihypertensive treatment and the prediction of major cardiovascular events. JAMA. 2004;292:

2343–2349.

19. Celik T, Iyisoy A, Kursaklioglu H, Kardesoglu E, Kilic S, Turhan H, Yilmaz MI, Ozcan O, Yaman H, Isik E, Fici F. Comparative effects of nebivolol and metoprolol on oxidative stress, insulin resistance, plasma adiponectin and soluble P-selectin levels in hypertensive patients.

J Hypertens. 2006;24:591–596.

20. Okin PM, Jern S, Devereux RB, Kjeldsen SE, Dahlof B; LIFE Study Group. Effect of obesity on electrocardiographic left ventricular hypertrophy in hypertensive patients: the Losartan Intervention for Endpoint (LIFE) Reduction in hypertension study. Hypertension.

2000;35:13–18.

21. Dahlof B, Devereux RB, Julius S, Kjeldsen SE, Beevers G, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Kristianson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H. Characteristics of 9194 patients with left ventricular hypertrophy: the LIFE study. Losartan Intervention For Endpoint Reduction in Hypertension. Hypertension. 1998;

32:989 –997.

22. Sokolow M, Lyon TP. The ventricular complex in left ventricular hyper- trophy as obtained by unipolar precordial and limb leads. 1949. Ann Noninvasive Electrocardiol. 2001;6:343–368.

23. Okin PM, Roman MJ, Devereux RB, Kligfield P. Electrocardiographic identification of increased left ventricular mass by simple voltage- duration products. J Am Coll Cardiol. 1995;25:417– 423.

24. McGavock JM, Victor RG, Unger RH, Szczepaniak LS; American College of Physicians and the American Physiological Society. Adiposity of the heart, revisited. Ann Intern Med. 2006;144:517–524.

25. Ciardullo AV, Azzolini L, Bevini M, Cadioli T, Daghio MM, Guidetti P, Lorenzetti M, Malavasi P, Morellini A, Carapezzi C. A diagnosis of left ventricular hypertrophy on ECG is associated with a high cardiovascular risk: findings from a 40- to 69-year-old cohort in general practice. Fam Pract. 2004;21:63– 65.

26. Huang Y, Walker KE, Hanley F, Narula J, Houser SR, Tulenko TN.

Cardiac systolic and diastolic dysfunction after a cholesterol-rich diet.

Circulation. 2004;109:97–102.

27. Liao Y, Takashima S, Maeda N, Ouchi N, Komamura K, Shimomura I, Hori M, Matsuzawa Y, Funahashi T, Kitakaze M. Exacerbation of heart failure in adiponectin-deficient mice due to impaired regulation of AMPK and glucose metabolism. Cardiovasc Res. 2005;67:705–713.

28. Rutter MK, Parise H, Benjamin EJ, Levy D, Larson MG, Meigs JB, Nesto RW, Wilson PW, Vasan RS. Impact of glucose intolerance and insulin resistance on cardiac structure and function: sex-related differences in the Framingham Heart Study. Circulation. 2003;107:448 – 454.

29. Murakami H, Ura N, Furuhashi M, Higashiura K, Miura T, Shimamoto K.

Role of adiponectin in insulin-resistant hypertension and atherosclerosis.

Hypertens Res. 2003;26:705–710.

30. Sega R, Trocino G, Lanzarotti A, Carugo S, Cesana G, Schiavina R, Valagussa F, Bombelli M, Giannattasio C, Zanchetti A, Mancia G. Alter- ations of cardiac structure in patients with isolated office, ambulatory, or home hypertension: data from the general population (Pressione Arteriose Monitorate E Loro Associazioni [PAMELA] Study). Circulation. 2001;

104:1385–1392.

31. Ryo M, Nakamura T, Kihara S, Kumada M, Shibazaki S, Takahashi M, Nagai M, Matsuzawa Y, Funahashi T. Adiponectin as a biomarker of the metabolic syndrome. Circ J. 2004;68:975–981.

32. Chan AY, Soltys CL, Young ME, Proud CG, Dyck JR. Activation of AMP-activated protein kinase inhibits protein synthesis associated with hypertrophy in the cardiac myocyte. J Biol Chem. 2004;279:

32771–32779.

33. Okin PM, Roman MJ, Devereux RB, Kligfield P. Electrocardiographic identification of left ventricular hypertrophy: test performance in relation to definition of hypertrophy and presence of obesity. J Am Coll Cardiol.

1996;27:124 –131.

34. Levy D, Labib SB, Anderson KM, Christiansen JC, Kannel WB, Castelli WP. Determinants of sensitivity and specificity of electrocardiographic criteria for left ventricular hypertrophy. Circulation. 1990;81:815– 820.

35. Abergel E, Tase M, Menard J, Chatellier G. Influence of obesity on the diagnostic value of electrocardiographic criteria for detecting left ventric- ular hypertrophy. Am J Cardiol. 1996;77:739 –744.

36. Alpert MA, Terry BE, Cohen MV, Fan TM, Painter JA, Massey CV. The electrocardiogram in morbid obesity. Am J Cardiol. 2000;85:908 –910, A10.

37. Okin PM, Roman MJ, Devereux RB, Kligfield P. ECG identification of left ventricular hypertrophy. Relationship of test performance to body habitus. J Electrocardiol. 1996;29 Suppl:256 –261.

38. Danias PG, Tritos NA, Stuber M, Kissinger KV, Salton CJ, Manning WJ. Cardiac structure and function in the obese: a cardiovascular magnetic resonance imaging study. J Cardiovasc Magn Reson. 2003;5:

431– 438.

39. Rao PS, Monarrez CN. Electrocardiographic differentiation of pos- terobasal left ventricular hypertrophy from right ventricular hypertrophy.

J Electrocardiol. 1981;14:25–30.

Downloaded from

(8)

40. The Examination Committee of Criteria for ‘Obesity Disease’ in Japan, Japan Society for the Study of Obesity. New criteria for ‘Obesity Disease’

in Japan. Circ J. 2002;66:987–992.

41. Tsugane S, Sasaki S, Tsubono Y. Under- and overweight impact on mortality among middle-aged Japanese men and women: a 10-y follow-up of JPHC study cohort I. Int J Obes Relat Metab Disord.

2002;26:529 –537.

42. Shiwaku K, Anuurad E, Enkhmaa B, Nogi A, Kitajima K, Shimono K, Yamane Y, Oyunsuren T. Overweight Japanese with body mass indexes of 23.0 –24.9 have higher risks for obesity-associated disorders: a com- parison of Japanese and Mongolians. Int J Obes Relat Metab Disord.

2004;28:152–158.

43. Vaccaro O, Cardoni O, Cuomo V, Panarelli W, Laurenzi M, Mancini M, Riccardi G, Zanchetti A; Gubbio Study Research Group. Rela- tionship between plasma insulin and left ventricular mass in normo- tensive participants of the Gubbio Study. Clin Endocrinol (Oxf).

2003;58:316 –322.

44. Smith J, Al-Amri M, Sniderman A, Cianflone K. Leptin and adiponectin in relation to body fat percentage, waist to hip ratio and the apoB/apoA1 ratio in Asian Indian and Caucasian men and women. Nutr Metab.

2006;3:18.

45. Smilde TD, Asselbergs FW, Hillege HL, Voors AA, Kors JA, Gansevoort RT, van Gilst WH, de Jong PE, Van Veldhuisen DJ. Mild renal dysfunction is associated with electrocardiographic left ventricular hypertro- phy. Am J Hypertens. 2005;18:342–347.

46. Isobe T, Saitoh S, Takagi S, Takeuchi H, Chiba Y, Katoh N, Shi- mamoto K. Influence of gender, age and renal function on plasma adiponectin level: the Tanno and Sobetsu study. Eur J Endocrinol.

2005;153:91–98.

47. Liu JE, Roman MJ, Pini R, Schwartz JE, Pickering TG, Devereux RB.

Cardiac and arterial target organ damage in adults with elevated ambu- latory and normal office blood pressure. Ann Intern Med. 1999;131:

564 –572.