ORIGINAL PAPER

Lower free testosterone level is correlated with left ventricular diastolic dysfunction in asymptomatic middle-aged men with type 2 diabetes mellitus Q. Jin,1 Y. Lou,2 H. Chen,1 T. Li,1 X. Bao,3 Q. Liu,1 X. He4

1

Department of Geriatrics, The Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, Zhejiang Province, China 2 Department of Clinical Laboratory, The First Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, Zhejiang Province, China 3 Department of Cardiac ultrasound, The Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, Zhejiang Province, China 4 Department of Emergency Medicine, The Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, Zhejiang Province, China Correspondence to: Xiaojun He, MD, Department of Emergency Medicine, The Second Affiliated Hospital of Zhejiang University School of Medicine, 88 Jiefang Road, Hangzhou, Zhejiang Province 310009, China Tel.: + 86 0571 87783951 Fax: + 86 0571 87783647 Email: [email protected]

Disclosures JQH, CHH, LTL, BXF, LQ and HXJ are full time employees of The Second Affiliated Hospital of Zhejiang University School of Medicine. LYF is full time employee of The First Affiliated Hospital of Zhejiang University School of Medicine. The authors have read the journal’s policies and have no conflicts of interest to declare.

1454

SUMMARY

What’s known

Aims: Low testosterone (T) level is associated with cardiovascular risk factors. However, the relationship between T level and heart function in asymptomatic men with type 2 diabetes mellitus (T2DM) is unknown. Methods: A total of 325 men were recruited who had no history, symptoms, or signs of heart disease. Results: T2DM had significantly lower free T (FT) levels than those with normal glucose metabolism (NGM) (428  38 pmol/l vs. 444  38 pmol/l, p = 0.0002), and had an increased risk of LVDD (66.1% vs. 31.8%). There was a significant difference in FT level between subjects with and without LVDD among those with T2DM (421  37 pmol/l vs. 442  40 pmol/l, p = 0.0007), but not among those with NGM (439  37 pmol/l vs. 447  39 pmol/l, p = 0.247) or in the group overall (426  38 pmol/l vs. 445  38 pmol/l, p = 0.156). Lower FT level was significantly associated with LVDD [univariate odds ratio (OR) = 0.63, p = 0.032; multivariate OR = 0.71, p = 0.039]. Receiver operating characteristic curve analysis of the usefulness of FT level for predicting LVDD showed an area under the curve (AUC) of 0.85 for T2DM (p < 0.001) and 0.66 for NGM (p < 0.05). FT level had a high predictive value for LVDD in T2DM (83% for FT < 414 pmol/l), but a low predictive value in NGM (61% for FT < 423 pmol/l). Comparison of the AUCs showed that FT level was more strongly correlated with LVDD in T2DM than in NGM. Conclusions: Lower FT level is correlated with LVDD in asymptomatic middle-aged men with T2DM.

Introduction Cardiovascular disease (CVD) is the primary cause of death in patients with diabetes mellitus (DM) (1). Male sex is an independent risk factor for CVD (2). Left ventricular (LV) diastolic dysfunction (LVDD) is recognised as an early manifestation of myocardial dysfunction in patients with type 2 DM (T2DM). LVDD has been demonstrated on echocardiography in 60–75% of asymptomatic T2DM patients (3,4). Previous population-based studies reported an increased risk of LVDD during the early stages of T2DM (4), and a 4- to 5-fold increased risk of HF in diabetic patients (5). Comorbidities such as hypertension, microangiopathy, myocardial ischaemia and renal dysfunction may explain the development of HF in patients with T2DM, but the prognosis of HF is worse in patients with T2DM than in those without (6). The associations between DM and diabetic

• • •

Cardiovascular disease is the primary cause of death in patients with diabetes mellitus. Low testosterone level is cardiovascular risk factors.

associated

with

Male sex is an independent risk factor for cardiovascular disease.

What’s new





Lower free testosterone level and asymptomatic left ventricular diastolic dysfunction were more prevalent and significant in men with type 2 diabetes. Lower free testosterone level is associated with left ventricular diastolic dysfunction in asymptomatic middle-aged men with type 2 diabetes.

cardiomyopathy, and the mechanisms underlying these associations, remain unclear. DM and low T level commonly coexist in men (7). T2DM is often associated with male hypogonadism (8). The relationship between androgen level and DM is important, as androgen level influence metabolism and the risk of vascular disease. In hypogonadal men with T2DM, there is some evidence that T supplementation have beneficial effects on important CVD risk factors including insulin resistance, glycemic control, lipid profile, central adiposity, body composition, and inflammatory state(8). Many studies found that, after adjustment for confounding variables, a low serum testosterone (T) level was associated with increased all-cause mortality that was independent of the risks associated with metabolic syndrome and DM (7). These results indicate that serum T level is closely related to DM and CVD.

ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461. doi: 10.1111/ijcp.12481

Testosterone and diastolic dysfunction

The role of T level in the development of CVD is complex and not completely understood. There is no information available regarding the relationships between T level and cardiac structure and function, and it is unknown whether T level affects the development of LVDD in men with T2DM. The aim of this study was to examine the relationships between serum T level and cardiac diastolic function in asymptomatic middle-aged men with T2DM. We hypothesised that a low T level, which nevertheless remains above the threshold for diagnosing hypogonadism, is a risk factor for LVDD in middle-aged men with T2DM who do not have HF.

Methods Patient population A total of 174 middle-aged men with T2DM were recruited from the inpatient department of the Division of Endocrinology at our hospital. The inclusion criteria were: (i) T2DM diagnosed according to the World Health Organization criteria: fasting plasma glucose level ≥ 126 mg/dl, or plasma glucose level ≥ 200 mg/dl at 2 h after a 75 g oral glucose load; (ii) Age 45–59 years; (iii) Taking oral antihyperglycaemic medication, but no insulin; (iv) LV ejection fraction (LVEF) ≥ 50% on echocardiography (9); (v) No indication of heart disease on history, physical examination or radiological investigation and (vi) Normal hepatic, renal, and pulmonary function and no major health problems other than T2DM. The exclusion criteria were: (i) Uncontrolled hypertension (blood pressure ≥ 180/110 mmHg), or HF with New York Heart Association functional class II–IV; (ii) Hormone replacement or glucocorticoid therapy; (iii) Hyperparathyroidism, thyroid dysfunction, or Cushing’s syndrome; (iv) Arrhythmia (frequent atrial or ventricular premature beats, tachycardia, atrial fibrillation or high-degree atrioventricular block); (v) Prostate cancer or other cancers; (vi) Serum total T (TT) level ≤ 12 nmol/l or free T (FT) level ≤ 225 pmol/l, and symptoms of low T level (erectile dysfunction, reduced muscle mass and strength, reduced energy, depressed mood) (10) and (vii) Taking medications that might affect serum T level such as opiates, cimetidine, glucocorticoids, phenytoin or spironolactone. As a control group, 151 middle-aged men with normal glucose metabolism (NGM) were recruited from the physical examination centre at our hospital. The study protocol was approved by the Medical Ethics Committee of the First Affiliated Hospital of Zhejiang University School of Medicine, and written informed consent was obtained from all subjects. ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461

Anthropometric, blood pressure, and laboratory data Blood samples were taken between 08:00 and 09:00 following an overnight fast. Measurements of haemoglobin A1c (HbA1c), glucose, C-peptide, insulin, lipids and other metabolic parameters were performed in the central laboratory of the hospital directly after obtaining the blood samples. Serum TT levels were measured by electrochemiluminescence (Siemens Immune 2000, Siemens Healthcare Diagnostics Products Limited, Shanghai, China) (gas chromatography/mass spectrometry is an ideal way). Sex hormone binding globulin (SHBG) levels were measured by ELISA (IBL, Inc., Hamburg, Germany). The intra- and interassay coefficients of variation for TT were 6.5% and 7.0%, respectively, and for SHBG were 5.5% and 9.0%, respectively. FT levels were calculated according to the formula of Vermeulen et al. (11) using an online calculator (http://www.issam.ch/freetesto.htm). For males aged 40–59 years, the reference range for TT was 13.2–30.8 nmol/l and for SHBG was 14.5–68.9 nmol/l. Serum brain natriuretic peptide (BNP) levels were measured by ELISA. In addition, the blood pressure, height and waist circumference of all subjects were measured. The homoeostasis model assessment of insulin resistance (HOMA-IR) was calculated using the formula [fasting insulin (mU/ ml) 9 fasting glucose (mmol/l)]/22.5.

Echocardiography Echocardiography is a simple, non-invasive modality for the assessment of cardiac function, and should be performed routinely in all diabetic patients before the development of cardiac symptoms. Transthoracic echocardiography was performed by a registered sonographer using standard equipment (Sonos 1000, Andover, MA, USA) and techniques, and was interpreted by an experienced cardiologist who was blinded to the clinical details. Assessment of LV diastolic function is an integral aspect of twodimensional Doppler and tissue Doppler echocardiography according to the guidelines and standards of the American Society of Echocardiography (12). Interventricular septal thickness (IVS), LV posterior wall thickness (LVPWT), and LV internal diameter (LVD) were measured at end-diastole (d) and end-systole. LV mass (LVM) and LVM index (LVMI) were calculated using the formulae recommended by the American Society of Echocardiography: LVM (g) = 0.8 9 (1.04 9 [(IVSd + LVPWTd + LVDd)3  LVDd3]) + 0.62, LVMI = LVM/height2.7. We used pulsed wave Doppler images to measure inflow at the tips of the mitral valve leaflets before (average of five measurements) and after the Valsalva manoeuvre, and calculated early-diastolic (E) and

1455

1456

Testosterone and diastolic dysfunction

late-diastolic (A) mitral flow velocities, deceleration time (DT) and isovolumetric relaxation time. We also used tissue Doppler images to measure peak early-diastolic mitral annulus velocity (e0 ) and peak atrial mitral annular diastolic velocity (a0 ). Normal LV diastolic function was defined as E/A 0.7–1.5, DT 140–240 ms, and e0 lateral ≥ 10 (13,14). LVDD was defined as an impaired, restrictive or pseudonormal relaxation pattern. An impaired relaxation pattern was defined as E/A < 0.7, DT > 240 ms, and e0 lateral < 10. A pseudonormal relaxation pattern was defined as E/A 0.7–1.5, DT 140–240 ms and e0 lateral < 10. A restrictive relaxation pattern was defined as E/A > 1.5, DT < 140 ms and e0 lateral < 10 (13,14).

Statistical analysis Continuous variables are expressed as mean  SD, and categorical variables are expressed as absolute number (percentage). All variables were tested for normality of distribution using the Kolmogorov– Smirnov test. Differences in continuous variables between groups were analysed using the unpaired ttest or Mann–Whitney U-test, as appropriate. The relationships between LVDD and the potential predictor variables were assessed using univariate logistic regression analysis. Independent predictors of LVDD were identified using multivariate logistic regression analysis that included an increased number of potential predictor variables for LVDD as well as factors that might affect heart function such as C-reactive protein (CRP) level, serum creatinine (Scr) level and urinary albumin excretion rate (UAER). The usefulness of FT level alone for predicting LVDD was analysed using receiver operating characteristic (ROC) curves. The results are expressed in terms of the area under the curve (AUC) and the 95% CI for the AUC, and AUCs were compared using the Z-test. A value of p < 0.05 was considered statistically significant. All analyses were performed using SPSS version 17.0 for Windows (SPSS, Chicago, IL, USA).

Results The study population consisted of 325 adults. Table 1 shows the median values of the anthropometric, blood pressure and laboratory data for all subjects. There were no significant differences between subjects with T2DM and with NGM in age or sex distribution, or in the factors indicating cardiovascular risk [heart rate, body mass index (BMI) and Scr level] (all p > 0.05). As expected, subjects with T2DM had significantly higher glucose and HbA1c levels, and higher HOMA-IR, than subjects with NGM. The median duration of disease in sub-

jects with T2DM was 5.4 years. There were no significant differences between subjects with T2DM and with NGM in frequency of hypertension, drug treatments, lipid levels, CRP level, Scr level, UAER or BNP level (all p > 0.05). TT and SHBG levels were slightly decreased in subjects with T2DM, but this was not significantly different compared with subjects with NGM (20.5  4.2 nmol/l vs. 21.3  4.1 nmol/l, p = 0.084; 37.5  5.1 nmol/l vs. 38.1  5.1 nmol/l, p = 0.291). FT level was significantly lower in subjects with T2DM than with NGM (428  38 pmol/l vs. 444  38 pmol/l, p = 0.0002). Echocardiographic parameters of cardiac structure and function in the two groups are shown in Table 2. There were no significant differences in cardiac structure between subjects with T2DM and with NGM. Subjects with T2DM had a longer DT, lower E/A and lower e0 than subjects with NGM, indicating that diastolic function was better in subjects with NGM than those with T2DM. Figure 1 shows the mean FT levels in subjects with LVDD (n = 163) and without LVDD (n = 162), which were not significantly different between these two groups (426  38 pmol/l vs. 445  38 pmol/l, p = 0.156). Among subjects with normal LVEF and cardiac structure, those with T2DM had a higher frequency of LVDD than those with NGM (66.1% vs. 31.8%, p < 0.05). Among subjects with T2DM, those with LVDD had lower FT levels than those without LVDD (421  37 pmol/l vs. 442  40 pmol/l, p = 0.0007). Among subjects with NGM, there was no significant difference in FT levels between those with and without LVDD (439  37 pmol/l vs. 447  39 pmol/l, p = 0.247). There was no significant difference in FT levels in normal LVDD with T2DM and NGM (442  40 pmol/l vs. 447  39 pmol/l, p = 0.438), but there was significant difference in FT levels in abnormal LVDD with T2DM and NGM (421  37 pmol/l vs. 439  37 pmol/l, p = 0.0052). Univariate logistic regression analysis was used to assess the relationships between LVDD and individual clinical and echocardiographic characteristics. As expected, diabetes, age, hypertension, HOMA-IR, HbA1c level, BNP level, FT level, left atrial diameter, LVRWT and LVMI were strongly associated with diastolic dysfunction. Higher FT level was associated with a lower risk of LVDD (OR = 0.63, p = 0.032) (Table 3). Multivariate logistic regression analysis was used to identify independent predictors of LVDD. The independent variables entered into the multivariate analysis were FT level and a larger number of potential predictor variables than used in the univariate analyses, and the dependent variable was LVDD. Age, T2DM, hypertension and BNP level ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461

Testosterone and diastolic dysfunction

Table 1 Baseline clinical parameters and testosterone levels in men with type 2 diabetes mellitus (T2GM) and with

normal glucose metabolism (NGM) Variable

T2DM (n = 174)

NGM (n = 151)

p

Age (years) HR (beat/min) Duration of diabetes (years) Height (m) Weight (kg) BMI (kg/m2) HbA1c (%) Albumin (g/l) FBG (mmol/l) 2hPG(mmol/l) HOMA-IR CRP(mg/dl) Scr (lmol/l) UAER (mg/24 h) TC (mmol/l) TG (mmol/l) HDL-C (mmol/l) LDL-C (mmol/l) DBP (mmHg) SBP(mmHg) TT (nmol/l) SHBG (nmol/l) FT (pmol/l) BNP (pg/ml) Hypertension (n, %) Antihypertensive (n, %)

52.4  78  5.4  1.67  67.8  24.31  7.72  40.5  7.34  9.87  5.28  2.52  87.41  132.11  2.74  2.13  1.11  3.25  77  154  20.5  37.5  428  106.44  72 (41.38) 72 (41.38)

52.1  77  – 1.68  67.6  23.92  5.63  40.7  5.23  5.84  4.97  2.41  83.63  125.33  2.63  2.05  1.14  3.18  75  150  21.3  38.1  444  101.76  56 (37.07) 56 (37.07)

0.379 0.426 – 0.262 0.819 0.195 < 0.001 0.627 < 0.001 < 0.001 0.010 0.076 0.139 0.055 0.079 0.141 0.223 0.501 0.223 0.147 0.084 0.291 0.0002 0.113 0.429 0.429

3.2 8 2.2 0.08 8.2 2.82 1.41 3.7 1.67 1.75 1.12 0.57 24.62 34.23 0.59 0.51 0.23 0.97 13 23 4.2 5.1 38 27.45

2.9 7 0.08 7.5 2.56 1.22 3.7 1.34 1.52 1.05 0.54 21.21 28.72 0.53 0.46 0.21 0.89 12 21 4.1 5.1 38 25.31

HR, heart rate; FBG, fasting blood glucose; 2hPG, 2-hour plasma glucose; HOMA-IR, homoeostasis model assessment of insulin resistance; CRP, C-reactive protein; Scr, serum creatinine; UAER, urinary albumin excretion rate; TC, total cholesterol; TG, triglyceride; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; DBP, diastolic blood pressure; SBP, systolic blood pressure; TT, total testosterone; FT, free testosterone; BNP, brain natriuretic peptide. Data are presented as mean  SD. Bold values indicate p < 0.05.

were identified as independent predictors of LVDD. FT level was the strongest independent predictor of LVDD (OR = 0.71, p = 0.039) (Table 4). The usefulness of FT level for predicting LVDD was analysed using ROC curve analysis (Figure 2). In subjects with T2DM, the AUC was 0.85 (95% CI 0.81–0.92, p < 0.001); an FT level of < 414 pmol/l had a sensitivity of 84%, specificity of 90%, accuracy of 88% and positive predictive value of 83% for predicting LVDD. In subjects with NGM, the AUC was only 0.66 (95% CI 0.62–0.79, p < 0.05); an FT level of < 423 pmol/l had a sensitivity of 76%, specificity of 81%, accuracy of 72% and positive predictive value of 61% for predicting LVDD. Comparison of the AUCs using the Z-test according to the formula z = |AUC1 AUC2|/√(SE12 + SE22) showed that FT level was more useful for predicting LVDD in subjects with T2DM than with NGM (z = 2.64, p < 0.01). ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461

Discussion Our study had two main findings. First, this study demonstrated lower FT levels in men with T2DM. Second, lower FT levels, which nevertheless remained in the normal range, were significantly associated with LVDD in men with T2DM, independent of traditional cardiovascular risk factors. While TT and SHBG levels were not associated with LVDD. This study clearly showed that the association between FT levels and markers of LV diastolic function was stronger in subjects with T2DM, indicating that FT level was a stronger predictor of LVDD in subjects with T2DM. In T2DM, diastolic dysfunction occurs before systolic dysfunction, and is a significant risk factor for HF (15). Studies reported high rates of LVDD in asymptomatic patients with T2DM who had normal systolic function (16). Early detection and treatment

1457

1458

Testosterone and diastolic dysfunction

Table 2 Baseline echocardiographic parameters in men with type 2 diabetes mellitus (T2GM) and with normal glucose

metabolism (NGM) T2DM (n = 174)

Characteristics

LV structure LVEDD (mm) LVESD (mm) LVEDV (mm3) LVESV (mm3) AV (mm) LAD (mm) LVRWT (mm) LVEF (%) LVM (g) LVMI (g/m2.7) LV function DT (ms) E (cm/s) A (cm/s) E/A e0 (cm/s) a0 (cm/s) E/e0

NGM (n = 151)

p

46.14 32.72 121.64 42.93 27.44 38.77 4.92 60.93 146.34 34.28

         

5.33 3.64 22.42 12.37 3.65 3.61 1.01 4.12 29.47 7.92

45.02 32.09 119.95 41.46 26.83 38.02 4.77 61.15 142.73 32.71

         

5.12 3.27 21.88 12.52 3.34 3.82 0.95 4.35 27.25 7.45

0.055 0.104 0.494 0.289 0.119 0.071 0.171 0.640 0.255 0.068

216.34 70.91 56.13 1.22 9.63 7.84 7.43

      

31.29 16.56 12.03 0.24 2.11 1.62 1.91

209.52 67.12 54.48 1.28 10.11 8.02 7.01

      

28.33 15.27 11.67 0.26 2.17 1.74 1.72

0.041 0.034 0.212 0.031 0.044 0.335 0.039

LV, left ventricular; EF, ejection fraction; LVEDD, LV end-diastolic diameter; LVESD, LV end-systolic diameter; LVEDV, LV end-diastolic volume; LVESV, LV end-systolic volume; AV, aortic diameter; LAD, left atrial diameter; LVRWT, LV relative wall thickness; LVM, LV mass; LVMI, LVM index; DT, deceleration time; E, early-diastolic filling velocity; A, late-diastolic filling velocity; e0 , early-diastolic mitral annulus velocity; a0 , atrial mitral annular diastolic velocity; E/e0 , LV end-diastolic pressure. Data are mean  SD. Bold values indicate p < 0.05.

(A)

(B)

Figure 1 Free testosterone (FT) levels in subjects with normal and abnormal left ventricular diastolic function. (A) FT levels in all subjects according to left ventricular diastolic function. (B) FT levels in subjects with type 2 diabetes mellitus (T2DM) and with normal glucose metabolism (NGM) according to left ventricular diastolic function. Data are expressed as meanSE. *vs. normal diastolic function in subjects with T2DM, p = 0.0052

of LV dysfunction can improve survival (17), and are especially important in patients with DM. Our finding showed that subjects with T2DM had a higher rate of LVDD than those with NGM, even though there was no significant difference in LV structure between the two groups. Many factors were known to be associated with LVDD, for example, DM (18), age (19), hypertension (19), BNP (20), HbA1c, insulin resistance (21) and CRP (22). DM with hypogonadotrophic hypogonadism has markedly elevated CRP concentrations (23). But there were no significant

changes of CRP levels in our study. All patients in our study had normal testosterone level and were under the influence of medications and dietary control (24), these may result in the lack of difference in CRP. Nevertheless, we found a stronger correlation between FT level and LVDD, univariate and multivariate analyses found that FT level was associated with LVDD (OR = 0.63 and 0.71, respectively, both p < 0.05). Approximately 50% of subjects in this study had LVDD. There were no significant difference in FT levels between those with LVDD and those without ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461

Testosterone and diastolic dysfunction

Table 3 Results of univariate logistic regression analyses

Table 4 Results of multivariate logistic regression

to evaluate the relationships between left ventricular diastolic dysfunction and potential predictor variables

analyses to evaluate the relationships between left ventricular diastolic dysfunction and potential predictor variables

Variables

Clinical Age DM Hypertension BMI HOMA-IR CRP TG TC LDL-c HDL-c SCr UAER HbA1c BNP TT SHBG FT Cardiac structure LVEDD LVESD LVEDV LVESV AV LAD LVRWT LVEF LVM LVMI

OR

95% CI

p

1.65 1.71 1.86 1.43 1.61 1.36 1.22 1.25 1.35 0.89 1.68 1.37 1.59 1.74 0.87 0.89 0.63

1.11–2.52 1.17–2.63 1.11–2.84 0.89–1.85 1.06–2.43 0.91–2.12 0.83–2.04 0.85–1.97 0.81–2.11 0.72–1.47 0.77–2.23 0.94–2.07 1.07–2.51 1.13–2.28 0.66–1.21 0.71–1.19 0.55–0.79

0.043 0.038 0.035 0.068 0.047 0.099 0.218 0.142 0.102 0.093 0.083 0.051 0.045 0.039 0.079 0.074 0.032

1.51 1.56 1.39 1.25 1.18 1.73 1.69 1.42 1.48 1.62

0.82–2.19 0.91–2.23 0.83–2.13 0.72–2.17 1.02–2.12 1.11–2.39 1.08–2.21 0.71–2.15 0.85–2.28 1.04–2.12

0.061 0.054 0.301 0.187 0.298 0.040 0.042 0.267 0.062 0.049

DM, diabetes mellitus; Scr, serum creatinine; HOMA-IR, homoeostasis model assessment of insulin resistance; CRP, Creactive protein; LDL-c, low-density lipoprotein cholesterol; TT, total testosterone; FT, free testosterone; UAER, urinary albumin excretion rate; LV, left ventricular; LVEF, LV ejection fraction; LVEDD, LV end-diastolic diameter; LVESD, LV end-systolic diameter; LVEDV, LV end-diastolic volume; LVESV, LV endsystolic volume; AV, aortic diameter; LAD, left atrial diameter; LVRWT, LV relative wall thickness; LVM, LV mass; LVMI, LVM index. Bold values indicate p < 0.05.

LVDD. However, in subjects with T2DM, FT levels were lower in those with LVDD than without LVDD (421  37 pmol/l vs. 442  40 pmol/l, p = 0.0007). The usefulness of FT level for predicting LVDD was assessed using ROC curve analysis. In subjects with T2DM, the AUC was 0.85, and an FT level of < 414 pmol/l had a sensitivity of 84%, specificity of 90% and accuracy of 88% for predicting LVDD. Men have a gradual decline in T level as they grow older (25), including men with T2DM. A previous study reported a high prevalence of low T level ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461

Age DM BMI Hypertension HbA1c CRP Scr LDL-c HDL-c UAER HOMA-IR BNP FT LAD LVRWT LVMI

OR

95% CI

p

1.42 1.54 1.32 1.51 1.34 1.25 1.36 1.22 0.91 1.31 1.42 1.57 0.71 1.48 1.35 1.28

1.04–2.73 1.09–2.81 0.78–2.02 1.05–2.92 1.02–2.74 0.73–2.27 0.67–2.38 0.73–2.27 0.68–2.32 0.85–2.24 0.91–2.72 1.11–2.57 0.51–0.84 1.04–2.52 0.91–2.33 0.96–2.27

0.048 0.045 0.074 0.044 0.057 0.137 0.177 0.276 0.172 0.143 0.077 0.044 0.039 0.042 0.123 0.058

DM, diabetes mellitus; Scr, serum creatinine; LDL-c, low-density lipoprotein cholesterol; HOMA-IR, homoeostasis model assessment of insulin resistance; CRP, C-reactive protein; FT, free testosterone; UAER, urinary albumin excretion rate; LAD, left atrial diameter; LVRWT, left ventricular relative wall thickness; LVMI, left ventricular mass index. Bold values indicate p < 0.05.

(23%) in men with T2DM (24). Men with T2DM are known to be at risk of low T levels and high HbA1c (26–28), decreased levels of FT were related to the severity of cardiac dysfunction in subjects with DM (29), which may explain why lower FT levels were more likely to be associated with LVDD in subjects with T2DM. There are many unanswered questions regarding the mechanisms by which T level may influence the development of LVDD. The proposed mechanism is that T modulates cardiovascular risk factors. This hypothesis is supported by the relationships between T level and high-density lipoprotein cholesterol level (30), blood pressure and ventricular mass (31), waist circumference (32), and HbA1c level (33). Intracoronary T infusion induces coronary vasodilatation and increases coronary blood flow in individuals with ischaemic heart disease who have low T levels (34). Higher T levels were found to be associated with a lower risk of cardiovascular events in elderly men (35), and were associated with better cardiac function through upregulation of major calcium regulatory proteins (36), and could be suppressed oxidative stress mediated by the androgen receptor-independent pathway (37).

1459

1460

Testosterone and diastolic dysfunction

LVDD in individuals with DM (39). However, such reductions in abnormal cardiac structure and function, and in the risk of CVD, were less significant in individuals with DM (40). It was not clear why T2DM had a greater risk of LVDD when the risk factors reached expected standards, and it was likely that the mechanism underlying this association was FT. Our findings suggest that lower FT level is significantly correlated with LVDD in asymptomatic men with T2DM. Measurement of FT levels could be used to identify asymptomatic patients with a normal LVEF who were at increased risk of developing LVDD. Considering the limited size of the sample in this study, further studies are needed to confirm the findings. Figure 2 Analysis of the usefulness of free testosterone (FT) level for predicting left ventricular diastolic dysfunction. Receiver operating characteristic curves comparing the sensitivity and specificity of FT level for predicting left ventricular diastolic dysfunction on echocardiography in subjects with type 2 diabetes mellitus (T2DM) and with normal glucose metabolism (NGM)

T2DM was a complex disease with multiple pathogenic factors, diabetic patients with LVDD were resistant to medical treatment (38). Intensive targeting of hyperglycaemia, hypertension, dyslipidaemia and lifestyle factors, aimed at protecting individuals from CVD, may have reduced the prevalence of

References 1 Morrish NJ, Wang SL, Stevens LK et al. 2001 Mortality and causes of death in the WHO Multinational Study of Vascular Disease in Diabetes. Diabetologia 2001; 44: S14–21. 2 Deverrux RB, Reichek N. Echocardiographic determination of left ventricular mass in man.Anatomic validation of the method. Circulation 1977; 55: 613–8. 3 Annonu AK, Fattah AA, Mokhtar MS et al. Left ventricular systolic and diastolic functional abnormalities in asymptomatic patients with non-insulin-dependent diabetes mellitus. J Am Soc Echocardiogr 2001; 14: 885–91. 4 Boyer JK, Thanigaraj S, Schechtman KB et al. Prevalence of ventricular diastolic dysfunction in asymptomatic, normotensive patients with diabetes mellitus. Am J Cardiol 2004; 93: 870–5. 5 Nichols GA, Gullion CM, Koro CE et al. The incidence of congestive heart failure in type 2 diabetes: an update. Diabetes Care 2004; 27: 1879–84. 6 Dries DL, Sweitzer NK, Drazner MH et al. Prognostic impact of diabetes mellitus in patients with heart failure according to the etiology of left ventricular systolic dysfunction. J Am Coll Cardiol 2001; 38: 421–8. 7 Laughlin GA, Barrett-Connor E, Bergstrom J. Low serum testosterone and mortality in older men. J Clin Endocrinol Metab 2008; 93: 68–75.

Acknowledgement This study was supported in part by the Department of Clinical Laboratory at the First Affiliated Hospital of Zhejiang University School of Medicine.

Author contributions JQH, LTL and CHH designed the research and wrote the article. LYF and BXF performed the research. HXJ and LQ contributed to analysing the data. All authors reviewed and approved the manuscript.

8 Corona G, Monami M, Rastrelli G et al. Type 2 diabetes mellitus and testosterone: a meta-analysis study. Int J Androl 2011; 34: 528–40. 9 Valle R, Bagolin E, Canali C et al. The BNP assay does not identify mild left ventricular diastolic dysfunction in asymptomatic diabetic patients. Eur J Echocardiogr 2006; 7: 40–4. 10 Lunenfeld B, Morley JE, Schulman C et al. Investigation, treatment and monitoring of late onset hypogonadism in males: ISA, ISSAM, and EAU recommendations. Int J Androl 2005; 28: 125–7. 11 Vermeulen A, Verdonck L. Kaufman JM A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999; 84: 3666–72. 12 Sahn DJ, DeMaria A, Kisslo J et al. Recommendations regarding quantifications in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978; 58: 1072–83. 13 Poulsen MK, Henriksen JE, Dahl J et al. Left ventricular diastolic function in type 2 diabetes mellitus: prevalence and association with myocardial and vascular disease. Circ Cardiovasc Imaging 2010; 3: 24–31. 14 Nagueh SF, Appleton CP, Gillebert TC et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr 2009; 22: 107–33. 15 Vanninen E, Mustonen J, Vainio P et al. Left ventricular function and dimensions in newly diag-

16

17

18

19

20

21

nosed non-insulin-dependent diabetes mellitus. Am J Cardiol 1992; 70: 371–8. Patil VC, Patil HV, Shah KB et al. Diastolic dysfunction in asymptomatic type 2 diabetes mellitus with normal systolic function. J Cardiovasc Dis Res 2011; 2: 213–22. From AM, Scott CG, Chen HH. Changes in diastolic dysfunction in diabetes mellitus over time. Am J Cardiol 2009; 103: 1463–6. Russo C, Jin Z, Homma S et al. Effect of diabetes and hypertension on left ventricular diastolic function in a high-risk population without evidence of heart disease. Eur J Heart Fail 2010; 12: 454–61. Fischer M, Baessler A, Hense HW et al. Prevalence of left ventricular diastolic dysfunction in the community. Results from a Doppler echocardiographic-based survey of a population sample. Eur Heart J 2003; 24: 320–8. van den Hurk K, Alssema M, Kamp O et al. Slightly elevated B-type natriuretic peptide levels in a non-heart failure range indicate a worse left ventricular diastolic function in individuals with, as compared with individuals without, type 2 diabetes: the Hoorn Study. Eur J Heart Fail 2010; 12: 958– 65. Giorda CB, Cioffi G, de Simone G et al. Predictors of early-stage left ventricular dysfunction in type 2 diabetes: results of DYDA study. Eur J Cardiovasc Prev Rehabil 2011; 18: 415–23.

ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461

Testosterone and diastolic dysfunction

22 Shah SJ, Marcus GM, Gerber IL et al. High-sensitivity C-reactive protein and parameters of left ventricular dysfunction. J Card Fail 2006; 12: 61–5. 23 Bhatia V, Chaudhuri A, Tomar R, Dhindsa S, Ghanim H, Dandona P. Low testosterone and high C-reactive protein concentrations predict low hematocrit in type 2 diabetes. Diabetes Care 2006; 29: 2289–94. 24 Pittas AG, Roberts SB, Das SK et al. The effects of the dietary glycemic load on type 2 diabetes risk factors during weight loss. Obesity 2006; 14: 2200– 9. 25 Hernandez-Mijares A, Garcıa-Malpartida K, Sola-Izquierdo E et al. Testosterone Levels in Males with Type 2 Diabetes and Their Relationship with Cardiovascular Risk Factors and Cardiovascular Disease. J Sex Med 2010; 7: 1954–64. 26 Ukkola O, Huttunen T, Puurunen VP et al. Total testosterone levels, metabolic parameters, cardiac remodeling and exercise capacity in coronary artery disease patients with different stages of glucose tolerance. Ann Med 2013; 45: 206–12. 27 Brand JS, Wareham NJ, Dowsett M et al. Associations of endogenous testosterone and SHBG with glycated haemoglobin in middle-aged and older men. Clin Endocrinol (Oxf) 2011; 74: 572–8. 28 Fukui M, Tanaka M, Hasegawa G et al. Association between serum bioavailable testosterone concentration and the ratio of glycated albumin to glycated hemoglobin in men with type 2 diabetes. Diabetes Care 2008; 31: 397–401.

ª 2014 John Wiley & Sons Ltd Int J Clin Pract, December 2014, 68, 12, 1454–1461

29 Araujo AB, O’Donnell AB, Brambilla DJ et al. Prevalence and incidence of androgen deficiency in middle-aged and older men: estimates from the Massachusetts Male Aging Study. J Clin Endocrinol Metab 2004; 89: 5920–6. 30 Haffner SM, Mykk€anen L, Valdez RA et al. Relationship of sex hormones to lipids and lipoproteins in nondiabetic men. J Clin Endocrinol Metab 1993; 77: 1610–5. 31 Svartberg J, von M€ uhlen D, Schirmer H et al. Association of endogenous testosterone with blood pressure and left ventricular mass in men. The Tromsø Study. Eur J Endocrinol 2004; 150: 65–71. 32 Svartberg J, von M€ uhlen D, Sundsfjord J et al. Waist circumference and testosterone levels in community dwelling men. The Tromsø Study. Eur J Epidemiol 2004; 19: 657–63. 33 Svartberg J, Jenssen T, Sundsfjord J et al. The associations of endogenous testosterone and sex hormone-binding globulin with glycosylated haemoglobin levels, in community dwelling men. The Ttomso Study. Diabetes Metab 2004; 30: 29– 34. 34 Webb CM, Elkington AG, Kraidly MM et al. Effects of oral testosterone treatment on myocardial perfusion and vascular function in men with low plasma testosterone and coronary heart disease. Am J Cardiol 2008; 101: 618–24. 35 Ohlsson C, Barrett-Connor E, Bhasin S et al. High serum testosterone is associated with reduced risk

36

37

38

39

40

1461

of cardiovascular events in elderly men. The MrOS (Osteoporotic Fractures in Men) study in Sweden. J Am Coll Cardiol 2011; 58: 1674–81. Gan L, Jiang W, Xiao YF et al. Disruption of CD38 gene enhances cardiac functions by elevating serum testosterone in the male null mice. Life Sci 2011; 89: 491–7. Zhang L, Wu S, Ruan Y et al. Testosterone suppresses oxidative stress via androgen receptor-independent pathway in murine cardiomyocytes. Mol Med Rep 2011; 4: 1183–8. Chen A, Wood M, Krauser D et al. NT-proBNP levels, echocardiographic findings, and outcomes in breathless patients: results from the ProBNP Investigation of Dyspnea in the Emergency Department (PRIDE) echocardiographic substudy. Eur Heart J 2006; 27: 839–45. Gaede P, Pedersen O. Intensive integrated therapy of type 2 diabetes: implications for long-term prognosis. Diabetes 2004; 53: S39–47. Gerdts E, Okin PM, Omvik P et al. Impact of diabetes on treatment induced changes in left ventricular structure and function in hypertensive patients with left ventricular hypertrophy. The LIFE study. Nutr Metab Cardiovasc Dis 2009; 19: 306–12.

Paper received November 2013, accepted May 2014

Lower free testosterone level is correlated with left ventricular diastolic dysfunction in asymptomatic middle-aged men with type 2 diabetes mellitus.

Low testosterone (T) level is associated with cardiovascular risk factors. However, the relationship between T level and heart function in asymptomati...
219KB Sizes 0 Downloads 4 Views