1. Unexplained cardiorespiratory arrests have been reported in patients with diabetic autonomic neuropathy and these could be due to denervation of the carotid chemoreceptors. 2. We have studied the ventilatory response to transient hypoxia ( ) during exercise in 22 male diabetic patients, six with symptomatic and cardiovascular evidence to suggest diabetic autonomic neuropathy (DAN+) and 12 without these features (DAN−). 3. There was no difference in the ventilatory response to transient hypoxia between the different groups of diabetic patients ( in DAN+ patients = −0·9 ± 0·2 litre min −1 kPa −1 ; in DAN− patients = −1·2 ± 0·6 litres min −1 kPa −1 ) even allowing for differences in the level of exercise achieved (CO 2 production in DAN+ patients = 743 ± 103 ml/min; CO 2 production in DAN– patients = 800 ± 144 ml/min). These results fell within our normal range for ventilatory response to transient hypoxia at this level of exercise. 4. The heart rate response to transient hypoxia varied within the groups but was significantly ( P < 0·05) less in the patients with established diabetic autonomic neuropathy. 5. We conclude that the peripheral chemoreceptors are intact in diabetic autonomic neuropathy and that other mechanisms must be implicated in the unexpected cardiorespiratory arrests seen in these patients. 1. Unexplained cardiorespiratory arrests have been reported in patients with diabetic autonomic neuropathy and these could be due to denervation of the carotid chemoreceptors. 2. We have studied the ventilatory response to transient hypoxia ( ) during exercise in 22 male diabetic patients, six with symptomatic and cardiovascular evidence to suggest diabetic autonomic neuropathy (DAN+) and 12 without these features (DAN−). 3. There was no difference in the ventilatory response to transient hypoxia between the different groups of diabetic patients ( in DAN+ patients = −0·9 ± 0·2 litre min −1 kPa −1 ; in DAN− patients = −1·2 ± 0·6 litres min −1 kPa −1 ) even allowing for differences in the level of exercise achieved (CO 2 production in DAN+ patients = 743 ± 103 ml/min; CO 2 production in DAN– patients = 800 ± 144 ml/min). These results fell within our normal range for ventilatory response to transient hypoxia at this level of exercise. 4. The heart rate response to transient hypoxia varied within the groups but was significantly ( P < 0·05) less in the patients with established diabetic autonomic neuropathy. 5. We conclude that the peripheral chemoreceptors are intact in diabetic autonomic neuropathy and that other mechanisms must be implicated in the unexpected cardiorespiratory arrests seen in these patients. 1. Unexplained cardiorespiratory arrests have been reported in patients with diabetic autonomic neuropathy and these could be due to denervation of the carotid chemoreceptors. 2. We have studied the ventilatory response to transient hypoxia ( ) during exercise in 22 male diabetic patients, six with symptomatic and cardiovascular evidence to suggest diabetic autonomic neuropathy (DAN+) and 12 without these features (DAN−). 3. There was no difference in the ventilatory response to transient hypoxia between the different groups of diabetic patients ( in DAN+ patients = −0·9 ± 0·2 litre min −1 kPa −1 ; in DAN− patients = −1·2 ± 0·6 litres min −1 kPa −1 ) even allowing for differences in the level of exercise achieved (CO 2 production in DAN+ patients = 743 ± 103 ml/min; CO 2 production in DAN– patients = 800 ± 144 ml/min). These results fell within our normal range for ventilatory response to transient hypoxia at this level of exercise. 4. The heart rate response to transient hypoxia varied within the groups but was significantly ( P < 0·05) less in the patients with established diabetic autonomic neuropathy. 5. We conclude that the peripheral chemoreceptors are intact in diabetic autonomic neuropathy and that other mechanisms must be implicated in the unexpected cardiorespiratory arrests seen in these patients.

1. Forced expiratory flow rates and specific airways conductance were measured before and after the inhalation of the atropine-like drug ipratropium bromide in 11 diabetic patients with autonomic neuropathy and a control group of 11 diabetic patients without autonomic neuropathy. Respiratory muscle function was assessed by recording maximal inspiratory and expiratory mouth pressures. 2. There was no difference between the two groups of diabetic patients in airway calibre before ipratropium bromide or in maximal mouth pressures. 3. Bronchodilatation after ipratropium bromide was greater in the control group than in the autonomic neuropathy group of patients. 4. It is concluded that, despite the lack of clinical evidence of respiratory involvement, patients with diabetic autonomic neuropathy have reduced airway vagal tone.

1. The heart-rate response during sustained hand grip was studied in four normal subjects before and after intravenous atropine, propranolol and combined cardiac autonomic blockade with both drugs. The results suggest that the increase in heart rate during the first 30 s is due to parasympathetic withdrawal, whereas the further increase between 30 s and 180 s is probably mediated by a combination of parasympathetic withdrawal and sympathetic stimulation. 2. The increases in heart rate during each minute of sustained hand grip were compared in 26 normal subjects, 37 diabetic subjects without and 24 diabetic subjects with proven autonomic neuropathy. In the diabetic subjects with autonomic neuropathy the increase in heart rate during the first minute was impaired, whereas the increases during the second and third minutes were similar in all three groups. 3. The initial increase in heart rate over the first 30 s of hand grip and the later increase between 30 s and 180 s were compared in nine normal subjects, ten diabetic subjects without and six diabetic subjects with autonomic neuropathy. The increase during the first 30 s was impaired in the diabetic subjects with autonomic neuropathy, whereas the later phase of the response was similar in all three groups. 4. It is concluded that impairment of the heart-rate response to sustained hand grip in diabetic autonomic neuropathy is mainly due to impairment of the early parasympathetic phase, but that the presence of cardiac sympathetic damage can also be detected.