Incidentally discovered adrenal masses are mostly benign, asymptomatic lesions, often arbitrarily considered as nonfunctioning tumors. Recent studies, however, have reported increasing evidence that subtle cortisol production and abnormalities in the hypothalamic-pituitary-adrenal (HPA) axis are more frequent than previously thought. The purpose of this study was to investigate the clinical and hormonal features of patients with incidentally discovered adrenal adenomas, in relation to their clinical outcome. Fifty consecutive patients with incidentally detected adrenal adenomas, selected from a total of 65 cases of adrenal incidentalomas, were prospectively evaluated. All of them underwent abdominal computed tomography scan and hormonal assays of the HPA axis function: circadian rhythm of plasma cortisol and ACTH, urinary cortisol excretion, 17-hydroxyprogesterone, androgens, corticotropin stimulation test and low-dose (2 mg) dexamethasone test. The patients were reevaluated at regular intervals (6, 12, and 24 months) for a median period of 38 months. Subtle hypercortisolism, defined as abnormal response to at least 2 standard tests of the HPA axis function in the absence of clinical signs of Cushing's syndrome (CS), was defined as subclinical CS. Mild-to-severe hypertension was found in 24 of 50 (48%) patients, type-2 diabetes in 12 of 50 (24%), and glucose intolerance in 6 of 50 (12%) patients. Moreover, 18 of 50 patients (36%) were diffusely obese (body mass index, determined as weight/height2, > 25), and 14 patients (28%) had serum lipid concentration abnormalities (cholesterol > or = 6.21 mmol/L, low-density lipoprotein cholesterol > or = 4.14 mmol/L and/or triglycerides > or = 1.8 mmol/L). Compared with a healthy population, bone mineral density Z-score, determined by the DEXA technique, tended to be slightly (but not significantly) lower in patients with adrenal adenoma (-0.41 SD). Endocrine data were compared with 107 sex- and age-matched controls, and patients with adenomas were found to have heterogeneous hormonal abnormalities. In particular, significantly higher serum cortisol values (P < 0.001), lower ACTH concentration (P < 0.05), and impaired cortisol suppression by dexamethasone (P < 0.001) were observed. Moreover, in patients with adenomas, cortisol, 17-OH progesterone, and androstenedione responses to corticotropin were significantly increased (P < 0.001, all), whereas dehydroepiandrosterone sulfate levels were significantly lower at baseline, with blunted response to corticotropin (P < 0.001, both). However, the criteria for subclinical CS were met by 12 of 50 (24%) patients. Of these, 6 (50%) were diffusely obese, 11 (91.6%) had mild-to-severe hypertension, 5 (41.6%) had type-2 diabetes mellitus, and 6 (50%) had abnormal serum lipids. The clinical and hormonal features improved in all patients treated by adrenalectomy, but seemed unchanged in all those who did not undergo surgery (follow-up, 9 to 73 months), except for one, who was previously found as having nonfunctioning adenoma and then revealed to have subclinical CS. In conclusion, an unexpectedly high prevalence of subtle autonomous cortisol secretion, associated with high occurrence of hypertension, diabetes mellitus, elevated lipids, and diffuse obesity, was found in incidentally discovered adrenal adenomas. Although the pathological entity of a subclinical hypercortisolism state remained mostly stable in time during follow-up, hypertension, metabolic disorders, and hormonal abnormalities improved in all patients treated by adrenalectomy. These findings support the hypothesis that clinically silent hypercortisolism is probably not completely asymptomatic.