In patients with congestive heart failure (CHF), sleep disordered breathing (SDB)--including obstructive and central sleep apnoea as well as periodic breathing--is a common condition and is believed to increase the risk of mortality. Treatment of SDB is considered important in the management of CHF. Improvements in SDB have a positive effect on cardiac output, measured with left ventricular ejection fraction (LVEF); on neurohormonal activity, measured as brain natriuretic peptide (BNP); and on the quality of life. Continuous positive airway pressure has been the traditional method used to treat SDB in patients with CHF, but compliance and tolerability are poor. A mandibular advancement device (MAD) is a dental device recommended for the treatment of sleep apnoea, but the method has never been evaluated in patients with CHF. The aims of the present studies were to evaluate the practical use of the MAD for the treatment of SDB in patients with CHF and to test the hypothesis that this intervention increases the dimensions of the pharyngeal airway (PAW), reduces SDB and BNP, and improves LVEF and the quality of life. Patients with mild to moderate CHF and SDB were evaluated using a portable polysomnographic device, lateral radiographs, cardiological and odontological examinations, and quality of life measures prior to and following intervention with an custom-made MAD. At the short-term follow-up 4-6 weeks after habituation with the MAD, the severity of SDB according to the apnoea-hypopnoea index had decreased from 25.1 +/- 9.4 (mean +/- SD) to 14.7 +/- 9.7 (p = 0.003). An increase in the inferior region of the PAW (7 +/- 5 mm) was observed on radiographs (p = 0.0001). However, no correlation between the effect of the MAD on the dimensions of the PAW and its effect on SDB was found. At the 6-month follow-up, the sleep apnoea-related symptoms had decreased by 31% (p = 0.003). Quality of life remained stable. BNP were reduced from 195.8 +/- 180.5 pg/ml to 148.1 +/- 139.9 pg/ml (p = 0.035). LVEF, however, remained unchanged. At the 12-month follow-up, 64 % of the patients were still using the MAD. Three patients withdrew from the study because of discomfort with the MAD. In most patients, MAD treatment had no severe side effect on the signs or symptoms of temporomandibular disorders. However, dental complications were observed. In conclusion, in patients with stable CHF who are experiencing problems with SDB, MAD intervention appears to reduce the severity of SDB, sleep apnoea-related symptoms, and neurohormonal activity. A lower tendency for PAW collapse may explain the effect observed on SDB. The reduction in plasma BNP may indicate decreased cardiac strain as a result of treatment of SDB. The 5-year survival rate, measured from the start of MAD intervention, was higher in the group that used a MAD than in the group that did not use a MAD (p = 0.036). No severe side effects on the stomatognathic system were observed during the intervention, and most patients--edentulous included--tolerated the treatment well. Impaired oral health, including reduced dentition and edentulousness, seemed to limit the use of the MAD in this group of elderly patients, both because of technical difficulties and because of the increased risk of dental complications. However, because the treatment of SDB is important in the management of CHF, the MAD intervention seems to be a valuable method in the treatment arsenal of SDB.