Effect of Eplerenone versus Spironolactone on Cortisol and Hemoglobin A1c Levels in Patients with Chronic Heart Failure

Masayuki Yamaji, MD; Takayoshi Tsutamoto, MD; Chiho Kawahara, MD; Keizo Nishiyama, MD; Takashi Yamamoto, MD; Masanori Fujii, MD; Minoru Horie, MD


Am Heart J. 2010;160(5):915-921. 

In This Article


In this study, we demonstrated for the first time that treatment with spironolactone for 4 months increased plasma HbA1c levels and serum cortisol levels in patients with CHF, while treatment with eplerenone did not. Furthermore, Δincrease in plasma HbA1c levels was positively correlated with Δincrease in cortisol levels after treatment with spironolactone. These findings suggest that eplerenone may be more useful than spironolactone for the treatment of CHF patients with DM, who easily progress to severe heart failure and show a poor prognosis.

Difference in the Effect on Cortisol Levels by Treatment with MR Antagonist

In the present study, plasma aldosterone levels were significantly increased by 4 months treatment with either spironolactone or eplerenone. Furthermore, serum cortisol levels were significantly increased by 4 months treatment with only spironolactone (Figure 1). Previously, it was reported that plasma aldosterone levels were increased by administration of both spironolactone and eplerenone and that plasma glucocorticoid levels were increased by administration of spironolactone in healthy subjects and diabetic patients,[31,33] but not eplerenone in Wistar rats,[34] consistent with our findings in mild CHF patients. However, the mechanism of this different effect between MR blockers remains to unclear.

MR antagonist binds and blocks not only MR but also other steroid hormone receptors such as glucocorticoid receptor and sex hormone receptor.[29] Since eplerenone is a spironolactone derivative designed to enhance selective binding to MR while minimizing binding to other steroid receptors, eplerenone and spironolactone show a differential effect in binding to glucocorticoid, androgen, and progesterone receptors, with binding affinities 100- to 1000-fold higher for spironolactone.[35,36] Hence, there is a possibility that the difference in the influence of these medicines on glucocorticoid receptor is related to the difference in the increase in cortisol levels.

Recently, cortisol levels were reported as an independent predictor of cardiac events in CHF patients.[17,18] Furthermore, it was recently reported that cardiac steroidogenesis was activated with an increase in cardiac glucocorticoid levels and glucocorticoid had the potential to induce cardiac hypertrophy via glucocorticoid receptor.[37] Thus, cortisol may have a harmful effect on patients with CHF and eplerenone may be superior to spironolactone in the treatment of CHF since it did not increase cortisol levels in the present study.

Difference in the Effect on HbA1c Levels by Treatment with MR Antagonist

The present study indicated that plasma HbA1c levels were significantly increased in patients receiving spironolactone, although those did not change in patients receiving eplerenone (Figure 1). In the CHARM study, it was indicated that administration of spironolactone may be associated with the development of DM as analyzed by univariate logistic regression analyses.[38] Furthermore, it has been reported that HbA1c levels were increased by treatment with spironolactone in patients with diabetes mellitus.[30,31] These findings are consistent with our results in mild CHF patients, but the mechanism of this effect remains to be clarified.

Elevated plasma cortisol levels, such as those in Cushing's syndrome or during treatment with therapeutic doses of steroid, often impair glucose tolerance and promote DM.[19,20] This process is mediated by impairment of insulin secretion from pancreatic islet cells mediated by glucocorticoid receptor, peripheral insulin resistance, and increased hepatic glucose output.[21,22,39] Swaminathan et al[31] reported that spironolactone significantly increased plasma HbA1c and cortisol in patients with diabetes mellitus. In the present study, serum cortisol levels were significantly increased by 4 months treatment with spironolactone, but not by that with eplerenone. In addition, the occupation and the inhibition of glucocorticoid receptor by MR antagonist may be insufficient. The differences in the effects on cortisol levels between these 2 MR antagonists might be related to these differences in metabolic effects, because the Δchange in HbA1c levels was positively correlated with Δchange in cortisol levels in the whole study population (r = 0.286, P = .003), especially in patients receiving spironolactone in the present study (Figure 2). Since HbA1c levels are a risk factor for mortality in CHF patients independent of DM,[10] eplerenone may be more useful than spironolactone considering the effects on both metabolism and heart failure, which mutually interact.

Difference in the Effect on Adiponectin Levels by Treatment with MR Antagonist

In the present study, plasma adiponectin levels were significantly decreased by treatment with spironolactone, but not treatment with eplerenone (Figure 1). Because plasma adiponectin levels increase with the severity of NYHA class function and correlate with plasma BNP levels,[28] plasma adiponectin levels would be decreased concomitant with improvement of CHF in patients receiving spironolactone. However, eplerenone also improved CHF since it decreased plasma BNP to the same levels as treatment with spironolactone, but plasma adiponectin levels did not change under eplerenone treatment. The mechanism underlying this differential effect on adiponectin between spironolactone and eplerenone remains unclear. However, Guo et al[40] reported that treatment of undifferentiated preadipocytes with aldosterone reduced mRNA and protein levels of peroxisome proliferator-activated receptor-γ and adiponectin and that eplerenone increased the expression of adiponectin in the heart and adipose tissue of obese, diabetic mice. The different effect of eplerenone on adiponectin levels may partly contribute to the selectivity to MR receptor compared with spironolactone although plasma BNP were decreased to the same levels by treatment with both MR antagonists. Further studies are necessary to clarify the differences in the effects of treatment with different MR antagonists on plasma adiponectin levels.

Study Limitations

Several limitations should be noted in the interpretation of these results. First, the small number of patients with CHF and the relatively short-term follow-up period and we did not measure or evaluate markers of glucose metabolism such as the fasting blood sugar or insulin levels. Further larger studies are needed to confirm the differences in the effects on glucose metabolism long term following treatment with different MR antagonists in patients with CHF.

Second the absolute changes in plasma adiponectin, HbA1c and cortisol levels were small and still within the normal range. However, it has been reported that normal circulating cortisol concentrations may be sufficient to predominantly activate cardiac MR in CHF and a higher cortisol levels is an independent predictor of cardiac event.[17,18] Further, HbA1c levels were reported to be an independent risk factor for mortality in the CHARM study population, which included many patients without diabetes and those showing a normal range of HbA1c levels.[10] Therefore, increases in HbA1c and cortisol levels in CHF patients may have clinical significance even though these increases are small and within the normal range. However, the present study only evaluated the changes in these variables based on 4 months treatment with MR antagonist, and did not follow the long-term prognosis and cardiac events. Further studies are needed to assess the long-term effects of MR antagonists on these levels and to evaluate the influence that changes in these variables exert on prognosis.

The interpretation of change in adiponectin levels is slightly confusing. High adiponectin levels in CHF patients are related to a poor prognosis.[27,28] However, plasma adiponectin levels are also negatively correlated with insulin resistance and insulin resistance increases the incidence of heart failure and is associated with a poor prognosis.[7–9,25,26] Therefore, it is unclear whether increased or decreased adiponectin levels are better for CHF patients, and further studies are necessary.